Control method for electric vehicle and control device for electric vehicle

Abstract

An electric vehicle includes a motor as a drive source and a coupling portion coupled to another vehicle, and travels while towing the coupled vehicle, which is the other vehicle coupled to the coupling portion. A control method for such an electric vehicle includes: calculating a basic torque target value representing a torque to be output by the motor based on a vehicle operation; calculating a final torque command value, which is a final command value for the torque, by performing correction processing for reducing a longitudinal vibration component generated in the electric vehicle due to the coupled vehicle being coupled to the coupling portion on the basic torque target value, based on a dynamic characteristic of the coupling portion to which the coupled vehicle is coupled; and controlling the motor based on the final torque command value.

Description

CROSS-REFERENCE TO RELATED APPLICATION(S)

This application claims priority to Japanese Patent Application Serial No. 2022-006018, filed Jan. 18, 2022, the entire disclosure of which is hereby incorporated by reference.

TECHNICAL FIELD

The present invention relates to a control method for an electric vehicle and a control device for an electric vehicle.

BACKGROUND

JP2003-009566A discloses a control method for reducing vibration in an electric vehicle based on a torque transmission characteristic of a power transmission mechanism connected between an output shaft of a motor and driving wheels.

SUMMARY

By vibration damping control according to a torque transmission characteristic of a power transmission mechanism, for example, vibration caused by a road surface gradient, a gear backlash, torsion of a drive shaft, and the like is reduced.

On the other hand, the electric vehicle may travel while towing the other vehicle or the like. When the electric vehicle travels while towing the other vehicle or the like, a total weight of the towing electric vehicle and the towed other vehicle is a substantial vehicle weight of the electric vehicle. Accordingly, when the electric vehicle performs towing traveling, a change in the vehicle weight to be considered is a large change that cannot be ignored.

In addition, since the torque transmission characteristic of the power transmission mechanism is modeled based on the weight of the electric vehicle, when there is a large change in the substantial vehicle weight due to towing the other vehicle or the like, a sufficient vibration damping effect may not be obtained by the vibration damping control in the related art. The torque transmission characteristic of the power transmission mechanism is usually determined by a structure of the power transmission mechanism or the like, and does not reflect the presence or absence of towing. Therefore, by the vibration damping control in the related art, vibration caused by towing the other vehicle or the like is not sufficiently reduced. That is, when the electric vehicle performs towing traveling, a sufficient vibration damping effect cannot be obtained by the vibration damping control in the related art, and vibration occurs in the electric vehicle. As a result, when performing towing traveling, the electric vehicle may not be able to achieve a rise in torque or smooth acceleration required by a vehicle operation.

An object of the present invention is to provide a control method for an electric vehicle and a control device for an electric vehicle capable of achieving a rise in torque and smooth acceleration required by a vehicle operation even when the electric vehicle performs towing traveling.

One aspect of the present invention provides a control method for an electric vehicle that includes a motor as a drive source and a coupling portion coupled to the other vehicle, and travels while towing a coupled vehicle, which is the other vehicle coupled to the coupling portion. In this control method, a basic torque target value representing a torque to be output by the motor is calculated based on the vehicle operation. A final torque command value, which is a final command value for the torque, is calculated by performing correction processing for reducing a longitudinal vibration component generated in the electric vehicle due to the coupled vehicle being coupled to the coupling portion on the basic torque target value, based on a dynamic characteristic of the coupling portion to which the coupled vehicle is coupled. Then, the motor is controlled based on this final torque command value.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is an explanatory diagram showing a schematic configuration of an electric vehicle;

FIG. 2 is a block diagram showing a configuration of a motor controller;

FIG. 3 is a graph showing an example of an accelerator opening-torque table;

FIG. 4 is an explanatory diagram showing a dynamic model of an electric vehicle in which a coupled vehicle is coupled to a coupling portion;

FIG. 5 is a block diagram showing a configuration of a longitudinal vibration damping unit;

FIG. 6 is a block diagram showing a configuration of a total weight calculation unit;

FIG. 7 is a block diagram showing a configuration of an electric vehicle weight calculation unit;

FIG. 8 is a block diagram showing a configuration of a natural vibration damping unit;

FIG. 9 is a block diagram showing a configuration of a power transmission mechanism vibration damping unit;

FIG. 10 is a graph showing a transmission characteristic H₂(s) used in a disturbance estimation unit;

FIG. 11 is a block diagram showing configurations of a current command value calculation unit and a current control processing unit;

FIGS. 12(A)-12(F) are a time chart showing a longitudinal acceleration and the like during towing traveling;

FIGS. 13(A)-13(F) are a time chart showing a longitudinal acceleration and the like when a weight of a coupled vehicle is large;

FIG. 14 is a block diagram showing a partial configuration of a vibration damping control unit in a second embodiment;

FIG. 15 is a block diagram showing a configuration of the longitudinal vibration damping unit in a third embodiment;

FIG. 16 is a block diagram showing a configuration of the natural vibration damping unit in the third embodiment;

FIG. 17 is a flowchart related to updating of a natural vibration frequency and the like;

FIGS. 18(A)-18(H) are a time chart showing the longitudinal acceleration and the like in a traveling scene in which the natural vibration frequency and the like are updated;

FIGS. 19(A)-19(H) are a time chart showing the longitudinal acceleration and the like after updating the natural vibration frequency and the like; and

FIG. 20 is a block diagram showing a partial configuration of the vibration damping control unit in a case where compensation for longitudinal vibration unique to towing traveling and compensation for power transmission mechanism vibration are substantially integrally performed.

DETAILED DESCRIPTION

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First Embodiment

FIG. 1 is an explanatory diagram showing a schematic configuration of an electric vehicle 100. The electric vehicle 100 is coupled to the other vehicle or the like, and can travel while towing the other vehicle or the like. The other vehicle or the like is, for example, a self-propelled vehicle or a vehicle (trailer) having no self-propelled capability. In the present embodiment, the electric vehicle 100 functions as a towing vehicle for towing the other vehicle or the like, but the electric vehicle 100 can travel without towing the other vehicle or the like. As shown in FIG. 1, the electric vehicle 100 includes a motor 10, a battery 11, an inverter 12, and a motor controller 13.

The motor 10 is a drive source of the electric vehicle 100. A torque generated by the motor 10 (hereinafter referred to as a motor torque T_m (not shown)) is transmitted to driving wheels 23 via a reduction gear 21 and a drive shaft 22. When the motor 10 is rotated by the driving wheels 23, the motor 10 can generate a regenerative braking force on the driving wheels 23 through so-called regenerative control. Then, the motor 10 recovers kinetic energy of the electric vehicle 100 as electric energy. In the present embodiment, the motor 10 is, for example, a three-phase AC synchronous motor. Currents i_u, i_v, and i_w flowing in the respective phases of the motor 10 can be detected using a current sensor 24. A rotor phase θ of the motor 10 is detected by a rotation sensor 25 such as a resolver or an encoder. Note that the reduction gear 21 and the drive shaft 22 constitute a power transmission mechanism (torque transmission system) that transmits power (torque) from the motor 10 to the driving wheels 23.

The battery 11 supplies electric power for driving the motor 10 via the inverter 12. The battery 11 can be charged with regenerative power generated in the motor 10 by the regenerative control. The battery 11 is a DC power supply. A DC voltage V_dc output from the battery 11 is detected by, for example, a voltage sensor 26, and is acquired directly from the voltage sensor 26 or via a battery controller (not shown).

The inverter 12 converts the DC power supplied from the battery 11 into AC power and supplies the AC power to the motor 10. The inverter 12 converts AC regenerative power input from the motor 10 by the regenerative control into DC power and inputs the DC power to the battery 11. The inverter 12 includes a plurality of switching elements (not shown), and converts DC power from the battery 11 into AC power by turning on and off the switching elements. Similarly, the inverter 12 converts AC regenerative power input from the motor 10 into DC power by turning on and off the switching elements. Two pairs of switching elements are provided for each phase of the motor 10. The switching elements are, for example, power semiconductor elements such as insulated gate bipolar transistors (IGBTs) or metal oxide film semi-conductor field-effect transistors (MOS-FET).

The motor controller 13 generates a PWM signal (pulse width modulation signal), which is a drive signal for the inverter 12, based on various vehicle variables, which are parameters indicating a control state of the electric vehicle 100 or each unit constituting the electric vehicle 100. Then, the motor controller 13 controls the motor 10 by driving the inverter 12 according to the generated PWM signal.

In the present embodiment, the motor controller 13 acquires or calculates, as the vehicle variables, the rotor phase θ, the currents i_u, i_v, i_w, the DC voltages V_dc, a longitudinal acceleration A_L1, an accelerator opening A_po, suspension stroke amounts ST_FL, ST_FR, ST_RL, ST_RR, a towing traveling signal SW_T, and the like.

The longitudinal acceleration A_L1 is an acceleration in a longitudinal direction (vertical direction) of the electric vehicle 100, and is detected by, for example, an acceleration sensor (not shown). The longitudinal acceleration A_L1 may be acquired from a controller (not shown). The accelerator opening A_po is a parameter indicating an operation amount of an accelerator provided in the electric vehicle 100, and is detected by a sensor (not shown) or the like. The accelerator opening A_po indicates a required amount of driving force (torque) for the electric vehicle 100 corresponding to a vehicle operation.

The suspension stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR are stroke amounts of suspensions provided on a left front wheel, a right front wheel, a left rear wheel, and a right rear wheel of the electric vehicle 100, respectively. The suspension stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR are detected by a suspension stroke sensor 27 provided in each suspension. Note that in the present embodiment, the front wheels of the electric vehicle 100 are the driving wheels 23, and the rear wheels are driven wheels.

The towing traveling signal SW_T is a signal output from a towing traveling switch 28. The towing traveling switch 28 is a control mode changeover switch that is operated by a driver or the like when the other vehicle is coupled to a coupling portion 101 (see FIG. 4) of the electric vehicle 100 and the electric vehicle 100 travels while towing the other vehicle. The coupling portion 101 is a member that is coupled to the other vehicle or the like, and is a towing member such as a so-called trailer hitch. Hereinafter, the other vehicle or the like (a vehicle to be towed or the like) coupled to the coupling portion 101 and towed by the electric vehicle 100 is referred to as a “coupled vehicle 102”.

In the present embodiment, the towing traveling signal SW_T is a determination criterion indicating whether the electric vehicle 100 is in towing traveling. Specifically, when the electric vehicle 100 travels while towing the coupled vehicle 102, the towing traveling signal SW_T is set to ON by an operation of the towing traveling switch 28. Then, the motor controller 13 determines whether the coupled vehicle 102 is coupled to the coupling portion 101, that is, whether the electric vehicle is in towing traveling, based on the towing traveling signal SW_T. Then, according to the determination result, the motor controller 13 changes a content of the vibration damping control for reducing the vibration generated in the electric vehicle 100.

FIG. 2 is a block diagram showing a configuration of the motor controller 13. The motor controller 13 is implemented by one or a plurality of computers, and is programmed so as to repeatedly execute processing of each unit described below in a predetermined cycle. Specifically, the motor controller 13 includes an input processing unit 31, a first torque target value calculation unit 32 (basic torque target value calculation unit), a vibration damping control unit 33 (correction processing unit), a current command value calculation unit 34, and a current control processing unit 35.

The input processing unit 31 acquires the vehicle variables used for various controls and/or calculations executed by the motor controller 13, or executes input processing for calculations. For example, the input processing unit 31 acquires, as the vehicle variables, the rotor phase θ, the currents i_u, i_v, the DC voltages V_dc, the longitudinal acceleration A_L1, the accelerator opening A_po, the suspension stroke amounts ST_FL, ST_FR, ST_RL, ST_RR, and the towing traveling signal SW_T.

The input processing unit 31 calculates a rotor angular velocity m [rad/s](electric angular velocity) of the motor 10 by differentiating the rotor phase θ (electric angle). The input processing unit 31 calculates a rotation speed of the motor 10 (hereinafter, referred to as a motor rotation speed) ω_m [rad/s](mechanical angular velocity) by dividing the rotor angular velocity ω by the number of pole pairs of the motor 10. In addition, the input processing unit 31 may calculate a motor rotation speed N_m [rpm] by multiplying the motor rotation speed ω_m by a unit conversion coefficient (60/2π).

In the present embodiment, the current sensor 24 detects the U-phase current i_u and the V-phase current i_v. Therefore, the input processing unit 31 calculates the W-phase current i_w using the U-phase current i_u and the V-phase current i_v based on the following equation (1).[Math. 1]i_w=−i_u−i_v  (1)

The first torque target value calculation unit 32 calculates a first torque target value T_m1* based on the accelerator opening A_po and the motor rotation speed ω_m. The first torque target value T_m1* is a torque target value calculated based on a vehicle operation, and indicates a torque (motor torque T_m) to be output by the motor 10. That is, the first torque target value T_m1* is a basic torque target value (basic torque target value) for generating the driving force required for the electric vehicle 100.

FIG. 3 is a graph showing an example of an accelerator opening-torque table. As shown in FIG. 3, it is a table in which the accelerator opening A_po and the motor rotation speed ω_m are associated with the first torque target value T_m1* in advance through experiment, simulation, or the like. The first torque target value calculation unit 32 calculates the first torque target value T_m1* by referring to the accelerator opening-torque table based on the accelerator opening A_po and the motor rotation speed ω_m.

The vibration damping control unit 33 (see FIG. 2) executes vibration damping control processing of calculating a sixth torque target value T_m6* by correcting the first torque target value T_m1* so as to reduce the vibration generated in the electric vehicle 100. The sixth torque target value T_m6* is used as a final torque command value, which is a final command value for the torque to be output by the motor 10.

In the present embodiment, there are two types of vibration generated in the electric vehicle 100. One is natural vibration that is generated by the coupled vehicle 102 being coupled to the coupling portion 101 when the electric vehicle 100 travels while towing the coupled vehicle 102. This vibration is generated in the longitudinal direction of the electric vehicle 100. Hereinafter, the vibration corresponding to the change in the acceleration in the longitudinal direction caused by the coupled vehicle 102 being coupled to the coupling portion 101 is referred to as longitudinal vibration. The other vibration is the vibration generated by transmitting the motor torque T_m to the driving wheels 23 by the power transmission mechanism. This vibration includes, for example, vibration caused by disturbance such as a road surface gradient, and vibration caused by gear backlash, torsion of the drive shaft 22, and the like. Hereinafter, the vibration generated in the power transmission mechanism is referred to as power transmission mechanism vibration.

In the present embodiment, in order to reduce the vibration, the vibration damping control unit 33 includes a longitudinal vibration damping unit 36 and a power transmission mechanism vibration damping unit 37.

The longitudinal vibration damping unit 36 executes longitudinal vibration damping processing. The longitudinal vibration damping processing is processing of reducing the longitudinal vibration generated when the coupled vehicle 102 is towed based on a dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled. The dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled refers to a change over time in movement of a coupling part (including the coupling portion 101) caused by the coupled vehicle 102 being coupled to the coupling portion 101. For example, a natural vibration frequency ω_t and an attenuation coefficient ζ_t are indices or parameters for identifying the change over time of the coupling part. Therefore, in the present embodiment, the change over time of the coupling part is identified by the natural vibration frequency ω_t, or the attenuation coefficient ζ_t, or both the natural vibration frequency ω_t and the attenuation coefficient ζ_t.

Specifically, the longitudinal vibration damping unit 36 determines whether the coupled vehicle 102 is coupled to the coupling portion 101 based on the towing traveling signal SW_T.

When the coupled vehicle 102 is coupled to the coupling portion 101, the longitudinal vibration damping unit 36 calculates a second torque target value T_m2* (see FIG. 5) by executing first vibration damping correction processing on the first torque target value T_m1* for reducing or eliminating a longitudinal vibration component. Then, the longitudinal vibration damping unit 36 outputs the second torque target value T_m2* as a third torque target value T_m3* output from the longitudinal vibration damping unit 36. The longitudinal vibration damping unit 36 performs the first vibration damping correction processing based on the longitudinal acceleration A_L1 and the suspension stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR.

On the other hand, when the coupled vehicle 102 is not coupled to the coupling portion 101, that is, when no longitudinal vibration is generated, the longitudinal vibration damping unit 36 outputs the first torque target value T_m1* as the third torque target value T_m3*.

The power transmission mechanism vibration damping unit 37 calculates the sixth torque target value T_m6* without sacrificing the response of the driving wheels 23 by performing second vibration damping correction processing (power transmission mechanism vibration damping processing) for further reducing the power transmission mechanism vibration on the third torque target value T_m3*. The power transmission mechanism vibration damping unit 37 performs the second vibration damping correction processing based on the motor rotation speed ω_m.

The current command value calculation unit 34 calculates a d-axis current target value i_d* and a q-axis current target value i_q* (hereinafter, referred to as dq-axis current target values i_d*, i_q*) based on the sixth torque target value T_m6*, the motor rotation speed ω_m, and the DC voltage V_dc. The current command value calculation unit 34 also calculates a d-axis non-interference voltage V_d-dcpl* and a q-axis non-interference voltage V_q-dcpl* (hereinafter, referred to as non-interference voltages V_d-dcpl*, V_q-dcpl*) in order to reduce a current due to interference between the d-axis and the q-axis. The current command value calculation unit 34 includes, for example, a map in which the sixth torque target value T_m6*, the motor rotation speed ω_m, and the DC voltage V_dc are associated with the dq-axis current target values i_d*, i_q* and the non-interference voltages V_d-dcpl*, V_q-dcpl* in advance by experiment or the like. Therefore, the current command value calculation unit 34 calculates the dq-axis current target values i_d*, i_q* and the non-interference voltages V_d-dcpl*, V_q-dcpl* corresponding to the sixth torque target value T_m6*, the motor rotation speed ω_m, and the DC voltage V_dc by referring to this map.

The current control processing unit 35 calculates, based on the dq-axis current target values i_d*, i_q* and the non-interference voltages V_d-dcpl*, V_q-dcpl*, PWM signals D_uu*, D_ul*, D_vu*, D_vl*, D_wu*, and D_wl* (see FIG. 11). The current control processing unit 35 drives the inverter 12 based on the PWM signals D_uu*, D_ul*, D_vu*, D_vl*, D_wu*, and D_wl*, and then outputs the motor torque T_m corresponding to the sixth torque target value T_m6*.

Hereinafter, specific configurations of the vibration damping control unit 33 and the current control processing unit 35 among the units of the motor controller 13 configured as described above will be described in detail.

[Configuration of Vibration Damping Control Unit]

First, a dynamic model of the electric vehicle 100 in which the coupled vehicle 102 is coupled to the coupling portion 101 and a transmission characteristic G_p(s) from the motor torque T_m to the motor rotation speed ω_m in the electric vehicle 100 will be described.

FIG. 4 is an explanatory diagram showing the dynamic model of the electric vehicle 100 in which the coupled vehicle 102 is coupled to the coupling portion 101. According to the dynamic model shown in FIG. 4, an equation of motion of the electric vehicle 100 can be expressed by the following equations (2) to (7). The equation of motion of the electric vehicle 100 in which the coupled vehicle 102 is coupled to the coupling portion 101 can be expressed by the following equations (8) to (11).

$\begin{array}{cc}\left[\mathrm{Math}.2\right]& \\ J_m·{\omega }_m·s=T_m-T_d/N& \left(2\right)\end{array}$ $\begin{array}{cc}2J_w·{\omega }_w·s=T_d-\mathrm{rF}& \left(3\right)\end{array}$ $\begin{array}{cc}{M}_1·V_1·s=F+F_r& \left(4\right)\end{array}$ $\begin{array}{cc}{T}_d=K_d·{\theta }_d& \left(5\right)\end{array}$ $\begin{array}{cc}F=K_t·\left(r{\omega }_m-v_1\right)& \left(6\right)\end{array}$ $\begin{array}{cc}{\theta }_d=\frac{{\theta }_m}{N}-{\theta }_w& \left(7\right)\end{array}$ $\begin{array}{cc}{F}_r=C_F·s·\left(x_2-x_1\right)+K_F·\left(x_2-x_1\right)& \left(8\right)\end{array}$ $\begin{array}{cc}{M}_2·s^2·x_1=-C_F,s·\left(x_2-x_1\right)-K_F·\left(x_2-x_1\right)& \left(9\right)\end{array}$ $\begin{array}{cc}{v}_1=s·x_1& \left(10\right)\end{array}$ $\begin{array}{cc}{v}_2=s·x_2& \left(11\right)\end{array}$

Parameters and the like shown in FIG. 4 and the above-described equations of motion are as follows. Note that “s” is a Laplace operator.

• • J_m: motor inertia
  • J_w: driving wheel inertia (for one wheel)
  • K_d: torsional rigidity of drive shaft
  • K_t: coefficient relating to tire and road surface
  • N: overall gear ratio
  • r: load radius of tire
  • ω_m: motor rotation speed
  • ω_w: driving wheel rotation speed
  • θ_m: motor angle
  • θ_w: driving wheel angle
  • T_m: motor torque
  • T_d: drive shaft torque
  • F: driving force (for two wheels)
  • v₁: vehicle speed of electric vehicle
  • v₂: vehicle speed of coupled vehicle
  • x₁: traveling distance of electric vehicle
  • x₂: traveling distance of coupled vehicle
  • M₁: weight of electric vehicle
  • M₂: weight of coupled vehicle
  • C_F: viscosity characteristic (viscosity coefficient)
  • K_F: elastic characteristic (elastic coefficient)
  • F_r: force applied to coupling portion

When the transmission characteristic G_p(s) from the motor torque T_m to the motor rotation speed ω_m is obtained based on the equations (2) to (7), the following equation (12) is obtained.

$\begin{array}{cc}\left[\mathrm{Math}.3\right]& \\ G_p\left(s\right)=\frac{1}{s}·\frac{b_3{s}^3+b_2{s}^2+b_{1}s+b_0}{a_3{s}^3+a_2{s}^2+a_{1}s+a_0}& \left(12\right)\end{array}$

Each parameter in the equation (12) is expressed by the equation (13) below. Note that in the equation (13), “M” is a total weight of the electric vehicle 100 and the coupled vehicle 102(=M₁+M₂).

$\begin{array}{cc}\left[\mathrm{Math}.4\right]& \\ \begin{array}{c}{a}_3=2J_m{J}_{w}M\\ a_2=K_t{J}_m\left(2J_w+r^{2}M\right)\\ a_1=K_{d}M\left(J_m+2J_w/N^2\right)\\ a_0=K_d{K}_t\left(J_m+2J_w/N^2+r^{2}M/N^2\right)\\ b_3=2J_{w}M\\ b_2=K_t\left(2J_w+r^{2}M\right)\\ b_1=2K_{d}M\\ b_0=2K_d{K}_t\end{array}& \left(13\right)\end{array}$

When a pole and a zero-point of the transmission characteristic G_p(s) shown in the equation (12) are examined, it can be approximated to the form shown in the following equation (14). Then, one pole and one zero-point show extremely close values. It means that α and β in the equation (14) indicate extremely close values.

$\begin{array}{cc}\left[\mathrm{Math}.5\right]& \\ G_p\left(s\right)=\frac{1}{s}·\frac{1\left(s+\beta \right)·\left(b_2^{\prime }{s}^2+b_1^{\prime }s+b_0^{\prime }\right)}{\left(s+\alpha \right)·\left(s^2+2{\zeta }_p{\omega }_{p}s+{\omega }_p^2\right)}& \left(14\right)\end{array}$

In accordance with the equation (14), by performing pole zero cancellation approximate to α=β, as shown in the following equation (15), a (second-order)/(third-order) transmission characteristic G_p(s) is obtained. In the equation (15), “ω_p” is a frequency of natural vibration generated by the transmission of the motor torque T_m by the power transmission mechanism. In the equation (15), “ζ_p” is an attenuation coefficient of the natural vibration generated by the power transmission mechanism.

$\begin{array}{cc}\left[\mathrm{Math}.6\right]& \\ G_p\left(s\right)=\frac{1}{s}·\frac{b_2^{\prime }{s}^2+b_1^{\prime }s+b_0^{\prime }}{s^2+2{\zeta }_p{\omega }_{p}s+{\omega }_p^2}& \left(15\right)\end{array}$

The transmission characteristic G_p(s) expressed by the above equation (15) is a vehicle model of the electric vehicle 100. A transmission characteristic G_r(s) of a model response when the attenuation coefficient ζ_p is set to “1” is expressed by the following equation (16).

$\begin{array}{cc}\left[\mathrm{Math}.7\right]& \\ G_r\left(s\right)=\frac{1}{s}·\frac{b_2^{\prime }{s}^2+b_1^{\prime }s+b_0^{\prime }}{s^2+2{\omega }_{p}s+{\omega }_p^2}& \left(16\right)\end{array}$

According to the equations (2) to (6) among the equations of motion of the electric vehicle 100, a transmission characteristic G_pF(s) from the motor torque T_m to a driving force F is expressed by the following equation (17).

$\begin{array}{cc}\left[\mathrm{Math}.8\right]& \\ G_{pF}\left(s\right)=\frac{c_0}{\left(s+\alpha \right)·\left(s^2+2{\zeta }_p{\omega }_{p}s+{\omega }_p^2\right)}& \left(17\right)\end{array}$

The transmission characteristic from the driving force F(s) of the electric vehicle 100 to the vehicle speed v₂(s) of the coupled vehicle 102 is expressed by the following equation (18) according to the equations (8) to (11) which are equations of motion of the electric vehicle 100 in which the coupled vehicle 102 is coupled to the coupling portion 101.

$\begin{array}{cc}\left[\mathrm{Math}.9\right]& \\ v_2\left(s\right)=\frac{C_{F}s+K_F}{s\left\{M_1{M}_2{s}^2+\left(M_1+M_2\right)C_{F}s+\left(M_1+M_2\right)K_F\right\}}F\left(s\right)& \left(18\right)\end{array}$

The transmission characteristic of the equation (18) can be approximated to a transmission characteristic of a secondary vibration system as shown in the following equation (19) by examining the poles thereof. In the equation (19), “ω_t” is a natural vibration frequency of the longitudinal vibration generated in the electric vehicle 100 due to the coupling of the coupled vehicle 102 to the coupling portion 101. In the equation (19), “ζ_t” is an attenuation coefficient of this natural vibration.

$\begin{array}{cc}\left[\mathrm{Math}.10\right]& \\ v_2\left(s\right)=\frac{1}{s}·\frac{C_{F}s+K_F}{\alpha \left(s^2+2{\zeta }_t{\omega }_{t}s+{\omega }_t^2\right)}F\left(s\right)& \left(19\right)\end{array}$

Hereinafter, specific configurations of the longitudinal vibration damping unit 36 and the power transmission mechanism vibration damping unit 37 constituting the vibration damping control unit 33 will be described in detail on the premise of the equations of motion, the transmission characteristics, and the like.

FIG. 5 is a block diagram showing the configuration of the longitudinal vibration damping unit 36 (see FIG. 2). As shown in FIG. 5, the longitudinal vibration damping unit 36 includes a total weight calculation unit 41, an electric vehicle weight calculation unit 42, a coupled vehicle weight calculation unit 43, a natural vibration damping unit 44, and a torque target value switching unit 45.

The total weight calculation unit 41 estimates a total weight M{circumflex over ( )} based on the longitudinal acceleration A_L1 and the third torque target value T_m3* (previous value). The total weight M{circumflex over ( )} is an estimation value of the total weight M that is the sum of the weight of the electric vehicle 100 (hereinafter referred to as an electric vehicle weight M₁) and the weight of the coupled vehicle 102 (hereinafter referred to as a coupled vehicle weight M₂). The total weight M{circumflex over ( )} is input to the coupled vehicle weight calculation unit 43.

The electric vehicle weight calculation unit 42 estimates the electric vehicle weight M₁ based on the suspension stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR. The electric vehicle weight M₁ is input to the coupled vehicle weight calculation unit 43 and the natural vibration damping unit 44.

The coupled vehicle weight calculation unit 43 estimates the coupled vehicle weight M₂ based on the total weight M{circumflex over ( )} and the electric vehicle weight M₁. In the present embodiment, the coupled vehicle weight calculation unit 43 calculates the coupled vehicle weight M₂ by subtracting the electric vehicle weight M₁ from the total weight M{circumflex over ( )}. The coupled vehicle weight M₂ is input to the natural vibration damping unit 44.

The natural vibration damping unit 44 performs the first vibration damping correction processing for reducing (lowering or eliminating) the longitudinal vibration component on the first torque target value T_m1* based on a dynamic characteristic of the coupling portion 101 coupled to the coupled vehicle 102. The natural vibration damping unit 44 identifies the dynamic characteristic (transmission characteristic) of the coupling portion 101 to which the coupled vehicle 102 is coupled based on the electric vehicle weight M₁, the coupled vehicle weight M₂, and a viscosity characteristic C_F, an elastic characteristic K_F, and the like, which are mechanical characteristics of the coupling portion 101. The dynamic characteristic of the coupling portion 101 corresponds to the natural longitudinal vibration generated in the electric vehicle 100 and the coupled vehicle 102.

Specifically, based on the electric vehicle weight M₁ and the coupled vehicle weight M₂, the natural vibration damping unit 44 calculates the second torque target value T_m2* by correcting the first torque target value T_m1* using the viscosity characteristic C_F and the elastic characteristic K_F. In this way, the natural vibration damping unit 44 reduces the longitudinal vibration, which is the natural vibration generated in the electric vehicle 100 during towing traveling. The second torque target value T_m2* is input to the torque target value switching unit 45.

The torque target value switching unit 45 determines whether the coupled vehicle 102 is coupled to the coupling portion 101 based on the towing traveling signal SW_T. Then, based on the determination result, the torque target value switching unit 45 switches the torque target value output as the third torque target value T_m3* to either the first torque target value T_m1* or the second torque target value T_m2*. Specifically, when the towing traveling signal SW_T is OFF and it is determined that the coupled vehicle 102 is not coupled, the torque target value switching unit 45 sets the first torque target value T_m1* as the third torque target value T_m3*, and outputs the third torque target value T_m3* to the power transmission mechanism vibration damping unit 37. On the other hand, when the towing traveling signal SW_T is ON and it is determined that the coupled vehicle 102 is coupled, the torque target value switching unit 45 sets the second torque target value T_m2* as the third torque target value T_m3* and outputs the third torque target value T_m3* to the power transmission mechanism vibration damping unit 37.

FIG. 6 is a block diagram showing a configuration of the total weight calculation unit 41 (see FIG. 5). As shown in FIG. 6, the total weight calculation unit 41 includes a first absolute value calculation unit 51, a second absolute value calculation unit 52, an acceleration difference calculation unit 53, a gain multiplication unit 54, an integration unit 55, a driving force calculation unit 56, and a longitudinal acceleration estimation unit 57.

The first absolute value calculation unit 51 calculates an absolute value |A_L1| of the longitudinal acceleration A_L1, which is a detection value. The absolute value |A_L1| of the longitudinal acceleration A_L is input to the acceleration difference calculation unit 53.

The second absolute value calculation unit 52 calculates an absolute value |A_L1{circumflex over ( )}| of a longitudinal acceleration estimation value A_L1{circumflex over ( )}. The longitudinal acceleration estimation value A_L1{circumflex over ( )} is an acceleration in the longitudinal direction estimated from the third torque target value T_m3*, and is calculated by the driving force calculation unit 56 and the longitudinal acceleration estimation unit 57 based on the third torque target value T_m3*. The absolute value |A_L1{circumflex over ( )}| of the longitudinal acceleration estimation value A_L1{circumflex over ( )} is input to the acceleration difference calculation unit 53.

The acceleration difference calculation unit 53 calculates an acceleration difference ΔA_L1 between the absolute value |A_L1| of the longitudinal acceleration A_L1 and the absolute value |A_L1{circumflex over ( )}| of the longitudinal acceleration estimation value A_L1{circumflex over ( )}. In the present embodiment, the acceleration difference ΔA_L1 is calculated by subtracting the absolute value |A_L1| of the longitudinal acceleration A_L1 from the absolute value |A_L1{circumflex over ( )}| of the longitudinal acceleration estimation value A_L1{circumflex over ( )}.

The gain multiplication unit 54 calculates a corrected total weight ΔM by multiplying the acceleration difference ΔA_L1 by a weight setting gain K_m. The weight setting gain Km is determined in advance by experiment, simulation, or the like. The corrected total weight ΔM represents a time change rate of the total weight M, that is, a change amount of the total weight M in each control cycle (unit time).

The integration unit 55 calculates the total weight M{circumflex over ( )} by adding up (integrating) the corrected total weight ΔM for each control cycle. Note that the integration unit 55 may initialize the total weight M{circumflex over ( )} based on a shift operation or the like by the driver. For example, the integration unit 55 acquires a shift operation signal S_shift. When the shift operation signal S_shift indicates a shift operation to a parking range, the integration unit 55 initializes the total weight M{circumflex over ( )}. An initial value of the total weight M{circumflex over ( )} is, for example, a design weight M_ini of the electric vehicle 100.

The driving force calculation unit 56 calculates the driving force F of the electric vehicle 100 based on the third torque target value T_m3*. In the present embodiment, the driving force F is calculated by applying the transmission characteristic G_pF(s) from the motor torque T_m to the driving force F to the third torque target value T_m3*.

The longitudinal acceleration estimation unit 57 estimates the longitudinal acceleration estimation value A_L1{circumflex over ( )} based on the driving force F calculated by the driving force calculation unit 56 and the total weight M{circumflex over ( )} (previous value). Specifically, the longitudinal acceleration estimation value A_L1{circumflex over ( )} is calculated by dividing the driving force F by the total weight M{circumflex over ( )}.

FIG. 7 is a block diagram showing a configuration of the electric vehicle weight calculation unit 42 (see FIG. 5). As shown in FIG. 7, the electric vehicle weight calculation unit 42 includes a total stroke amount calculation unit 58, a stroke amount change calculation unit 59, a weight change calculation unit 60, and an addition unit 61.

The total stroke amount calculation unit 58 calculates a total stroke amount EST of each suspension by adding up the suspension stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR.

The stroke amount change calculation unit 59 calculates a suspension stroke change amount ΔST by subtracting the total stroke amount ΣST from a reference total stroke amount ST_ini. The reference total stroke amount ST_ini is a reference value determined in advance by a design of the electric vehicle 100.

The weight change calculation unit 60 calculates a weight change amount ΔM₁ of electric vehicle 100 by multiplying the suspension stroke change amount ΔST by a spring constant K_ST [N/mm]. The spring constant K_ST is determined in advance according to the design of the electric vehicle 100 (each suspension).

The addition unit 61 calculates the electric vehicle weight M₁, which is an estimation value of the weight of the electric vehicle 100, by adding the design weight M_ini of the electric vehicle 100 and the weight change amount ΔM₁.

FIG. 8 is a block diagram showing a configuration of the natural vibration damping unit 44 (see FIG. 5). As illustrated in FIG. 8, the natural vibration damping unit 44 includes a feature amount calculation unit 62 and a first vibration damping correction processing unit 63.

The feature amount calculation unit 62 identifies the dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled by calculating feature amounts of the longitudinal vibration caused by the coupled vehicle 102 being coupled to the coupling portion 101. Specifically, the feature amount calculation unit 62 calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t, which are the feature amounts of the longitudinal vibration, based on the electric vehicle weight M₁ and the coupled vehicle weight M₂, according to the above-described equations (18) and (19). As shown in the equations (18) and (19), the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101 are used to calculate the natural vibration frequency ω_t and the attenuation coefficient ζ_t of the longitudinal vibration. The coupling portion 101 is, for example, a genuine part of the electric vehicle 100. Therefore, in the present embodiment, the viscosity characteristic C_F and the elastic characteristic K_F are known according to a design of the coupling portion 101. The natural vibration frequency ω_t and the attenuation coefficient ζ_t of the longitudinal vibration are input to the first vibration damping correction processing unit 63.

The first vibration damping correction processing unit 63 calculates the second torque target value T_m2* by correcting the first torque target value T_m1* based on the natural vibration frequency ω_t and the attenuation coefficient ζ_t of the longitudinal vibration. The correction processing performed by the first vibration damping correction processing unit 63 is the first vibration damping correction processing for lowering or eliminating the longitudinal vibration component.

The first vibration damping correction processing unit 63 is implemented by, for example, a band-stop filter in which a frequency band (center frequency) for reducing a signal and a gain in the frequency band are variable. The first vibration damping correction processing unit 63 sets the frequency band (center frequency) of the band-stop filter according to the natural vibration frequency ω_t of the longitudinal vibration. In the present embodiment, as shown in FIG. 8, the center frequency is set to the natural vibration frequency ω_t of the longitudinal vibration. The first vibration damping correction processing unit 63 sets the gain in the center frequency of the band-stop filter according to the attenuation coefficient ζ_t of the longitudinal vibration. In the present embodiment, as shown in FIG. 8, the gain for the center frequency (the natural vibration frequency ω_t of the longitudinal vibration) is set to be smaller as the attenuation coefficient ζ_t of the longitudinal vibration is smaller. That is, it is set so that the smaller the attenuation coefficient ζ_t is and the less the attenuation of the longitudinal vibration is, the larger a drop of the gain in the center frequency ω_t is. Accordingly, the longitudinal vibration is efficiently reduced.

More specifically, the first vibration damping correction processing unit 63 is implemented by, for example, a notch filter, which is one form of the band-stop filter. A transmission characteristic of the notch filter is expressed by the following equation (20). In the equation (20), “ω” is a center frequency, and “ζ” is an attenuation coefficient. “D” is a parameter (hereinafter, simply referred to as a gain D) that defines a drop depth of the gain in the center frequency. Therefore, the first vibration damping correction processing unit 63 sets the center frequency ω of the notch filter to the natural vibration frequency ω_t of the longitudinal vibration. The first vibration damping correction processing unit 63 sets the gain D of the notch filter in accordance with the attenuation coefficient ζ_t of the longitudinal vibration. Note that the attenuation coefficient ζ of the notch filter is related to a width of the frequency band in which the gain is reduced. The attenuation coefficient ζ_t of the notch filter is determined in advance based on experiment, simulation, or the like, and is preferably set to at least a value of 1 or more. Even when a band-stop filter other than the notch filter is used, setting thereof is the same as the setting of the notch filter.

$\begin{array}{cc}\left[\mathrm{Math}.11\right]& \\ \frac{s^2+D·2\mathrm{\zeta \omega }s+{\omega }^2}{s^2+2\mathrm{\zeta \omega }s+{\omega }^2}& \left(20\right)\end{array}$

FIG. 9 is a block diagram showing a configuration of the power transmission mechanism vibration damping unit 37 (see FIG. 2). As shown in FIG. 9, the power transmission mechanism vibration damping unit 37 includes a feedforward compensation unit 64, a feedback compensation unit 65, and an addition unit 66.

The feedforward compensation unit 64 compensates in advance the power transmission mechanism vibration caused by transmitting the motor torque T_m corresponding to the third torque target value T_m3* to the driving wheels 23 through the power transmission mechanism, using the transmission characteristic G_p(s) which is the vehicle model of the electric vehicle 100. That is, the feedforward compensation unit 64 calculates a fourth torque target value T_m4* by compensating for a component of the power transmission mechanism vibration included in the third torque target value T_m3*. The feedforward compensation unit 64 is expressed by a transmission characteristic G_r(s)/G_p(s) constituted by the transmission characteristic G_p(s) which is the vehicle model and a transmission characteristic G_r(s) of a model response. The fourth torque target value T_m4*, which is a torque target value after the feedforward compensation, is input to the addition unit 66.

The feedback compensation unit 65 compensates for the power transmission mechanism vibration caused by disturbance such as a road surface gradient based on the sixth torque target value T_m6* (previous value) serving as the final torque command value and the actual motor rotation speed ω_m. Specifically, the feedback compensation unit 65 includes a motor rotation speed estimation unit 71, a deviation calculation unit 72, a disturbance estimation unit 73, and a gain multiplication unit 74.

The motor rotation speed estimation unit 71 calculates a motor rotation speed estimation value ω_m{circumflex over ( )} which is an estimation value of the motor rotation speed ω_m from the sixth torque target value T_m6* by using the transmission characteristic G_p(s) which is the vehicle model of the electric vehicle 100.

The deviation calculation unit 72 calculates a deviation (hereinafter, referred to as a motor rotation speed deviation Δω_m) between the motor rotation speed ω_m, which is a detection value, and the motor rotation speed estimation value ω_m{circumflex over ( )}. In the present embodiment, the motor rotation speed deviation Δω_m is calculated by subtracting the motor rotation speed ω_m from the motor rotation speed estimation value ω_m{circumflex over ( )}.

The disturbance estimation unit 73 calculates a disturbance estimation value d{circumflex over ( )} based on the motor rotation speed deviation Δω_m. The disturbance estimation value d{circumflex over ( )} is an estimation value for the disturbance such as a road surface gradient. The disturbance estimation unit 73 is expressed by, for example, a transmission characteristics H₂(s)/G_p(s). A transmission characteristic H₂(s) is set such that a difference between a denominator order and a numerator order thereof is equal to or larger than a difference between a denominator order and a numerator order of the transmission characteristic G_p(s). The disturbance estimation value d{circumflex over ( )} is input to the gain multiplication unit 74.

The gain multiplication unit 74 calculates a fifth torque target value T_m5* by multiplying the disturbance estimation value d{circumflex over ( )} by a feedback gain K_FB. The feedback gain K_FB is determined in advance by, for example, experiment, simulation, or the like. The fifth torque target value T_m5* represents a compensation amount according to the disturbance with respect to the torque related to the power transmission mechanism vibration caused by the disturbance, that is, the motor torque T_m. The fifth torque target value T_m5* is input to the addition unit 66.

The addition unit 66 calculates the sixth torque target value T_m6* by adding the fourth torque target value T_m4*, which is a torque target value after the feedforward compensation, to the fifth torque target value T_m5*, which is a feedback torque. As described above, the sixth torque target value T_m6* is a final command value for the motor torque T_m.

As described above, the second vibration damping correction processing performed by the power transmission mechanism vibration damping unit 37 includes compensation processing by the feedforward compensation unit 64 and compensation processing by the feedback compensation unit 65.

Note that FIG. 10 is a graph showing the transmission characteristic H₂(s) used in the disturbance estimation unit 73. As shown in FIG. 10, the transmission characteristic H₂(s) is, for example, a band-pass filter. When the transmission characteristic H₂(s) is a band-pass filter, the feedback compensation unit 65 is a feedback element that selectively reduces a vibration component due to the disturbance. The transmission characteristic H₂(s) is configured such that, for example, an attenuation coefficient on a low-pass side and an attenuation coefficient on a high-pass side substantially coincide with each other. The transmission characteristic H₂(s) is set such that the center frequency f_p of the passing frequency band thereof substantially coincides with a torsional resonance frequency ω_p of the power transmission mechanism (particularly, the drive shaft 22). Note that the horizontal axis (frequency) in FIG. 10 is a logarithmic scale. Especially, when the transmission characteristic H₂(s) includes a first-order high-pass filter and a first-order low-pass filter, the transmission characteristic H₂(s) is expressed by the following equation (21).

$\begin{array}{cc}\left[\mathrm{Math}.12\right]& \\ H_2\left(s\right)=\frac{{\tau }_{H}s}{\left(1+{\tau }_{H}s\right)·\left(1+{\tau }_{L}s\right)}& \left(21\right)\end{array}$

In the equation (21), “τ_H” and “τ_L” are a time constant of the high-pass filter and a time constant of the low-pass filter, respectively. τ_L=1/(2πf_HC), f_HC=k·f_p, τ_H=1/(2πf_LC), and f_LC=f_p/k. Note that “k” is any constant, and “f_HC” and “f_LC” are cutoff frequencies on the high-frequency side and the low-frequency side, respectively.

[Configuration of Current Control Processing Unit]

FIG. 11 is a block diagram showing the configuration of the current control processing unit 35 (see FIG. 2). As shown in FIG. 11, the current control processing unit 35 includes a voltage command value calculation unit 81, a coordinate conversion unit 82, a PWM conversion unit 83, and a coordinate conversion unit 84.

The voltage command value calculation unit 81 calculates smoothed non-interference voltages V_d-dcpl-flt*, V_q-dcpl-flt* by processing the non-interference voltages V_d-dcpl*, V_q-dcpl* with a low-pass filter. Then, the voltage command value calculation unit 81 calculates a d-axis voltage command value V_d* and a q-axis voltage command value V_q* (hereinafter referred to as dq-axis voltage command values V_d*, V_q*) by so-called current control calculation based on the dq-axis currents i_d, i_q, the dq-axis current target values i_d*, i_q*, and the smoothed non-interference voltages V_d-dcpl-flt*, V_q-dcpl-flt*. The dq-axis currents i_d, i_q are calculated by the coordinate conversion unit 84.

The coordinate conversion unit 82 converts the dq-axis voltage command values V_d*, V_q* into voltage command values of U, V, and W phases (hereinafter, referred to as three-phase voltage command values V_u*, V_v*, and V_w*) based on the rotor phase θ of the motor 10 according to the following equation (22).

$\begin{array}{cc}\left[\mathrm{Math}.13\right]& \\ \left[\begin{array}{c}{V}_u^{*}\\ V_v^{*}\\ V_w^{*}\end{array}\right]=\sqrt{\frac{2}{3}}\left[\begin{array}{cc}1& 0\\ -\frac{1}{2}& \frac{\sqrt{3}}{2}\\ -\frac{1}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{cc}\cos \theta & -\sin \theta \\ \sin \theta & \cos \theta \end{array}\right]\left[\begin{array}{c}{V}_d^{*}\\ V_q^{*}\end{array}\right]& \left(22\right)\end{array}$

The PWM conversion unit 83 generates the PWM signals D_uu*, D_ul*, D_vu*, D_vl*, D_wu*, and D_wl*, which are drive signals of the switching elements in the inverter 12, according to the three-phase voltage command values V_u*, V_v*, and V_w*. When the inverter 12 is driven according to the PWM signals, the motor 10 is controlled to output the motor torque T_m according to the sixth torque target value T_m6*.

The coordinate conversion unit 84 calculates the W-phase current i_w based on the U-phase current i_u and the V-phase current i_v, detected by the current sensor 24. Then, the coordinate conversion unit 84 converts the currents i_u, i_v, and i_w into the dq-axis currents i_d, i_q using the rotor phase θ of the motor 10 according to the following equation (23). The dq-axis currents i_d, i_q are used by the voltage command value calculation unit 81 as described above.

$\begin{array}{cc}\left[\mathrm{Math}.14\right]& \\ \left[\begin{array}{c}{i}_d\\ i_q\end{array}\right]=\left[\begin{array}{cc}\cos \theta & \sin \theta \\ -\sin \theta & \cos \theta \end{array}\right]\sqrt{\frac{2}{3}}\left[\begin{array}{ccc}1& -\frac{1}{2}& -\frac{1}{2}\\ 0& \frac{\sqrt{3}}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{c}{i}_u\\ i_v\\ i_w\end{array}\right]& \left(23\right)\end{array}$

Note that the input processing unit 31 includes a motor rotation speed calculation unit 85. The motor rotation speed calculation unit 85 calculates the motor rotation speed ω_m based on the rotor phase θ of the motor 10.

[Effects]

Hereinafter, effects in a case where the electric vehicle 100 configured as described above including the coupling portion 101 coupled to the coupled vehicle 102 travels while towing the coupled vehicle 102 will be described.

FIGS. 12(A)-12(F) are a time chart showing the longitudinal acceleration A_L1 and the like during towing traveling. Specifically, in FIG. 12(A) is a time chart of the first torque target value T_m1*. In FIG. 12(B) is a time chart of the sixth torque target value T_m6*. In FIG. 12(C) is a time chart of the drive shaft torque T_d. In FIG. 12(D) is a time chart of the motor rotation speed ω_m. In FIG. 12(E) is a time chart showing the longitudinal acceleration A_L1 of the electric vehicle 100. In FIG. 12(F) is a time chart showing the longitudinal acceleration A_L2 of the coupled vehicle 102.

In FIGS. 12(A)-12(F), a solid line indicates an example of a case where the vibration damping control according to the present embodiment is performed, and a broken line indicates an example of a case where vibration damping control according to a comparative example is performed. The comparative example is an example in which the longitudinal vibration damping unit 36 of the vibration damping control unit 33 does not perform the first vibration damping correction processing. However, in the comparative example, the second vibration damping correction processing is still performed by the power transmission mechanism vibration damping unit 37. Therefore, the sixth torque target value T_m6* of the comparative example shown in FIG. 12(B) is the same value as the sixth torque target value T_m6* in a case where the coupled vehicle 102 is not towed in the present embodiment (in a case where T_m3*=T_m1*).

As shown in FIG. 12(A), here, it is assumed that the first torque target value T_m1* is input in a step at a time t₁ by an accelerator operation of the driver. As shown in FIG. 12(B), after the time t₁, at a time t₂, the sixth torque target value T_m6* increases to a predetermined torque corresponding to a request by the accelerator operation (accelerator opening A_po). As shown in FIG. 12(C), the drive shaft torque T_d changes in accordance with the sixth torque target value T_m6*. However, in both the present embodiment (solid line) and the comparative example (broken line), the power transmission mechanism vibration is compensated by the second vibration damping correction processing, and therefore, the change in the drive shaft torque T_d is smooth. As shown in FIG. 12(D), the motor rotation speed ω_m changes in accordance with the change in the sixth torque target value T_m6*.

On the other hand, a difference appears in the longitudinal acceleration A_L1 of the electric vehicle 100 and the longitudinal acceleration A_L2 of the coupled vehicle 102 between the vibration damping control according to the present embodiment and the vibration damping control of the comparative example. Specifically, as shown in FIG. 12(E) and FIG. 12(F), in the control (broken line) of the comparative example in which the first vibration damping correction processing for reducing the longitudinal vibration due to the towing traveling is not performed, the longitudinal vibration is generated between the electric vehicle 100 and the coupled vehicle 102 due to towing the coupled vehicle 102. As a result, in the control of the comparative example, vibration appears in the longitudinal accelerations A_L1 and A_L2 of the electric vehicle 100 and the coupled vehicle 102. The longitudinal vibration continues even after the time t₂ when the sixth torque target value T_m6* converges to a torque corresponding to the accelerator opening A_po. Especially, this longitudinal vibration continues even at a time t₃ when a sufficient time elapses after the sixth torque target value T_m6* converges. It indicates that the longitudinal vibration due to towing traveling cannot be sufficiently reduced only by the second vibration damping correction processing for compensating for the power transmission mechanism vibration. In contrast, in the control (solid line) of the present embodiment, the first vibration damping correction processing is performed, and thus the longitudinal vibration between the electric vehicle 100 and the coupled vehicle 102 is reduced. Especially, as shown in FIG. 12(E), even in a transition period (from the time t₁ to the time t₂) in which the sixth torque target value T_m6* converges to the torque corresponding to the accelerator opening A_po, a sufficient vibration damping effect of reducing the longitudinal vibration appears. In the control (solid line) of the present embodiment, as shown in FIG. 12(F), the longitudinal vibration of the electric vehicle 100 towing the coupled vehicle 102 is reduced, and the longitudinal vibration of the coupled vehicle 102 which is the towed vehicle is also reduced. As a result, even in the case of towing traveling, the torque required by the vehicle operation can be increased and smooth acceleration can be achieved.

FIGS. 13(A)-13(F) are a time chart showing the longitudinal acceleration A_L1 and the like when the weight of the coupled vehicle 102 is large. That is, FIGS. 13(A)-13(F) show a case where the weight M₂ of the coupled vehicle 102 is larger than that in the case of FIGS. 12(A)-12(F). Parameters shown in FIGS. 13(A)-13(F), differences between the solid line and the broken line, and the like are the same as those in FIGS. 12(A)-12(F).

As shown in FIGS. 13(A)-13(C), even when the weight M₂ of the coupled vehicle 102 is large, the changes in the first torque target value T_m1*, the sixth torque target value T_m6*, and the drive shaft torque T_d are the same as those in FIGS. 12(A) and 12(B). As shown in FIG. 13(D), an increase in the motor rotation speed ω_m is reduced by an increase in the weight M₂ of the coupled vehicle 102, but the increase tendency is the same as that shown in FIG. 12(D).

As shown in FIG. 13(E) and FIG. 13(F), in the control (broken line) of the comparative example, the amplitudes of the vibrations appearing in the longitudinal accelerations A_L1 and A_L2 of the electric vehicle 100 and the coupled vehicle 102 are increased as compared with the cases shown in FIG. 12(E) and FIG. 12(F). Frequencies of the vibration appearing in the longitudinal accelerations A_L1 and A_L2 are small. Further, the attenuation of the longitudinal vibration is also delayed, and a large longitudinal vibration continues after the time t₃. These changes are caused by a change in the characteristic of the longitudinal vibration appearing in the electric vehicle 100 and the coupled vehicle 102, that is, the dynamic characteristics (the natural vibration frequency ω_t and the attenuation coefficient ζ_t) of the coupling portion 101 to which the coupled vehicle 102 is coupled, which is caused by the increase in the weight M₂ of the coupled vehicle 102.

In contrast, in the control (solid line) according to the present embodiment, since the weight M₂ of the coupled vehicle 102 is increased, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 are accurately reduced even when the dynamic characteristics are changed as described above.

As described above, in the case where the electric vehicle 100 travels while towing the coupled vehicle 102, unique longitudinal vibration during towing traveling is generated as compared with a case where the electric vehicle 100 travels alone. This longitudinal vibration is caused by a large change in the weight (total weight M{circumflex over ( )}) of the entire system and a substantial change in the transmission characteristic of the electric vehicle 100. This longitudinal vibration is not sufficiently reduced only by the vibration damping control (the second vibration damping correction processing) for compensating for the power transmission mechanism vibration.

In the control according to the present embodiment, the first torque target value T_m1* is corrected by the first vibration damping correction processing to calculate the sixth torque target value T_m6* as the final torque command value, and the motor 10 is controlled according to the sixth torque target value T_m6*. Accordingly, the unique longitudinal vibration in towing traveling is reduced. In the control according to the present embodiment, the effect of reducing the longitudinal vibration is obtained regardless of the weight M₂ of the coupled vehicle 102 which is the towed vehicle. Accordingly, as a result, in the case of towing traveling, regardless of the weight M₂ of the coupled vehicle 102, the increase of the torque required by the vehicle operation and the smooth acceleration are achieved.

Second Embodiment

In the first embodiment, the vibration damping control unit 33 includes the longitudinal vibration damping unit 36 and the power transmission mechanism vibration damping unit 37, and the compensation for the longitudinal vibration unique to the towing traveling and the compensation for the power transmission mechanism vibration are separately performed. For example, the vibration damping control unit 33 may be configured to substantially integrally perform the compensation for the longitudinal vibration unique to the towing traveling and the compensation for the power transmission mechanism vibration. Hereinafter, a second embodiment in which the longitudinal vibration damping unit 36 and the feedforward compensation unit 64 (see FIG. 9) which is a part of the power transmission mechanism vibration damping unit 37 are integrally configured will be described as an example.

FIG. 14 is a block diagram showing a partial configuration of the vibration damping control unit 33 in the second embodiment. As shown in FIG. 14, the vibration damping control unit 33 of the second embodiment does not explicitly include the longitudinal vibration damping unit 36 of the first embodiment. Therefore, in the vibration damping control unit 33 of the second embodiment, the feedforward compensation unit 64 is configured to directly calculate the fourth torque target value T_m4* based on the first torque target value T_m1*.

This feedforward compensation unit 64 of the second embodiment includes a vehicle model of the electric vehicle 100 (hereinafter referred to as an electric vehicle model 91), a model of the coupled vehicle 102 coupled to the coupling portion 101 (hereinafter referred to as a coupled vehicle model 92), a compensation torque calculation unit 93, and a vibration damping correction processing unit 94.

The electric vehicle model 91 calculates an estimation value of the motor torque T_m corresponding to the fourth torque target value T_m4* (hereinafter, simply referred to as a torque estimation value T_m{circumflex over ( )}) based on the fourth torque target value T_m4* (previous value) output from the feedforward compensation unit 64. Specifically, the electric vehicle model 91 is basically configured in accordance with the equations (2) to (7), which are equations of motion of the electric vehicle 100. The electric vehicle model 91 is configured to calculate the vehicle speed v₁ of the electric vehicle 100 using a force f₁₂ acting between the electric vehicle 100 and the coupled vehicle 102 based on the equations (8) to (11). In the present embodiment, the force f₁₂ is calculated by the coupled vehicle model 92. The weight M₁ of the electric vehicle 100 used in the calculation processing of the vehicle speed v₁ of the electric vehicle 100 is calculated by the same configuration as that of the electric vehicle weight calculation unit 42 of the first embodiment. That is, the weight M₁ of the electric vehicle 100 used in the electric vehicle model 91 is a variable parameter. The torque estimation value T_m{circumflex over ( )} is input to the compensation torque calculation unit 93. The vehicle speed v₁ is input to the coupled vehicle model 92.

The coupled vehicle model 92 is configured to calculate the vehicle speed v₂ of the coupled vehicle 102 according to the equations (8) to (11). Specifically, the vehicle speed v₂ of the coupled vehicle 102 is calculated based on a deviation between the vehicle speed v₁ of the electric vehicle 100 and the vehicle speed v₂ of the coupled vehicle 102, that is, a relative vehicle speed Δv₁₂ between the electric vehicle 100 and the coupled vehicle 102. In the calculation of the relative vehicle speed Δv₁₂, the vehicle speed v₁ of the electric vehicle 100 is acquired from the electric vehicle model 91, and a previous value of a value calculated by the coupled vehicle model 92 is used as the vehicle speed v₂ of the coupled vehicle 102.

The viscosity characteristic C_F and the elastic characteristic K_F, which are the mechanical characteristics of the coupling portion 101, are used to calculate the vehicle speed v₂ of the coupled vehicle 102. Accordingly, the dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled is substantially reflected in the value of the vehicle speed v₂ of the coupled vehicle 102. The weight M₂ of the coupled vehicle 102 used in the calculation processing the vehicle speed v₂ of the coupled vehicle 102 is calculated by the same configurations as the electric vehicle weight calculation unit 42, the total weight calculation unit 41, and the coupled vehicle weight calculation unit 43 of the first embodiment. That is, the weight M₂ of the coupled vehicle 102 used in the coupled vehicle model 92 is a variable parameter. The force f₁₂ calculated in the calculation processing of the vehicle speed v₂ is input to the electric vehicle model 91. The relative vehicle speed Δv₁₂ is input to the compensation torque calculation unit 93. Note that the relative vehicle speed Δv₁₂ substantially reflects the dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled via the vehicle speed v₂ of the coupled vehicle 102.

The compensation torque calculation unit 93 calculates a first compensation torque T_FF1* by multiplying the torque estimation value T_m{circumflex over ( )} by a predetermined gain K₁ determined in advance by experiment, simulation, or the like. The first compensation torque T_FF1* compensates for the torque corresponding to the power transmission mechanism vibration in advance.

The compensation torque calculation unit 93 calculates a second compensation torque T_FF2* based on the relative vehicle speed Δv₁₂. The second compensation torque T_FF2* compensates in advance for the torque corresponding to the longitudinal vibration caused by the coupled vehicle 102 being coupled to the coupling portion 101. More specifically, when the coupled vehicle 102 is coupled to the coupling portion 101, the compensation torque calculation unit 93 sets a value obtained by multiplying the relative vehicle speed Δv₁₂ by a predetermined gain K₂ determined in advance by experiment, simulation, or the like as the second compensation torque T_FF2*. On the other hand, when the coupled vehicle 102 is not coupled to the coupling portion 101, the compensation torque calculation unit 93 sets zero (“0”), which is a fixed value, as the second compensation torque T_FF2*.

The switching of the second compensation torque T_FF2* is performed by a second compensation torque switching unit 95 based on the towing traveling signal SW_T. That is, when the second compensation torque switching unit 95 determines that the coupled vehicle 102 is coupled to the coupling portion 101 based on the towing traveling signal SW_T, the second compensation torque T_FF2* becomes a value based on the relative vehicle speed Δv₁₂. On the other hand, when the second compensation torque switching unit 95 determines that the coupled vehicle 102 is not coupled to the coupling portion 101 based on the towing traveling signal SW_T, the second compensation torque T_FF2* is set to zero.

The compensation torque calculation unit 93 calculates a third compensation torque T_FF3* by adding up the first compensation torque T_FF1* and the second compensation torque T_FF2* calculated as described above. Therefore, the third compensation torque T_FF3* is, in principle, a torque for compensating in advance for the power transmission mechanism vibration, but when the coupled vehicle 102 is coupled to the coupling portion 101, the third compensation torque T_FF3* further includes a torque for compensating in advance for the longitudinal vibration unique to the towing traveling. The third compensation torque T_FF3* is input to the vibration damping correction processing unit 94.

The vibration damping correction processing unit 94 calculates the fourth torque target value T_m4* by correcting the first torque target value T_m1* using the third compensation torque T_FF3*. When the coupled vehicle 102 is coupled to the coupling portion 101, the correction processing performed by the vibration damping correction processing unit 94 substantially includes the first vibration damping correction processing and the correction processing including the feedforward compensation of the first embodiment. Therefore, the fourth torque target value T_m4* calculated by the vibration damping correction processing unit 94 is substantially the same as the fourth torque target value T_m4* output by the feedforward compensation unit 64 of the first embodiment. Therefore, similarly to the first embodiment, the fourth torque target value T_m4* becomes the sixth torque target value T_m6* by the feedback compensation unit 65 and the addition unit 66, and is used as the final torque command value.

As described above, the longitudinal vibration damping unit 36 and the feedforward compensation unit 64 of the first embodiment can be integrally configured. Since the fourth torque target value T_m4* output by the feedforward compensation unit 64 of the second embodiment in which the above two units are integrated is substantially the same as the fourth torque target value T_m4* of the first embodiment, as described above, even when the longitudinal vibration damping unit 36 and the feedforward compensation unit 64 of the first embodiment are integrated, the same functions and effects as those of the first embodiment can be achieved (see FIGS. 12(A)-(F) and 13(A)-(F)).

Note that the feedforward compensation unit 64 of the second embodiment substantially performs the vibration damping control based on the transmission characteristic from the input of the first torque target value T_m1* to the output of the fourth torque target value T_m4*. The electric vehicle weight M₁ and the coupled vehicle weight M₂ constituting the transmission characteristic are variable parameters, and the content of the second compensation torque T_FF2* is switched depending on the presence or absence of the coupled vehicle 102. Accordingly, the control according to the second embodiment is configured to reduce the longitudinal vibration unique to the towing traveling by changing (adjusting) the content of the transmission characteristic depending on whether the coupled vehicle 102 is coupled to the coupling portion 101.

Third Embodiment

In the first embodiment and the second embodiment, it is assumed that the elastic characteristic K_F and the viscosity characteristic C_F, which are the mechanical characteristics of the coupling portion 101, are known, and an error (modeling error) included in the transmission characteristic G_p(s), which is the vehicle model of the electric vehicle 100, is sufficiently small to an ignorable extent. However, more practically, errors of the elastic characteristic K_F and the viscosity characteristic C_F of the coupling portion 101 may not be ignorable when, for example, a trailer hitch manufactured by a third party is used. The elastic characteristic K_F or the viscosity characteristic C_F of the coupling portion 101 may change due to a change over time, and an error may occur. Depending on an actual situation of towing traveling, a modeling error of the transmission characteristic G_p(s) and the like may not be ignored. In such a situation, even if the correction processing for reducing the longitudinal vibration in the first embodiment or the second embodiment is performed, the longitudinal vibration may still be generated. In the present third embodiment, a configuration in which the longitudinal vibration can be suitably reduced regardless of an error included in the known values such as the elastic characteristic K_F and the viscosity characteristic C_F of the coupling portion 101 or a modeling error will be described.

FIG. 15 is a block diagram showing a configuration of the longitudinal vibration damping unit 36 in the third embodiment. As shown in FIG. 15, in the longitudinal vibration damping unit 36 of the present third embodiment, the longitudinal acceleration A_L1 is input to the natural vibration damping unit 44. The natural vibration damping unit 44 performs the first vibration damping correction processing on the first torque target value T_m1* using the longitudinal acceleration A_L1 in addition to the electric vehicle weight M₁, the coupled vehicle weight M₂, and the viscosity characteristic C_F and the elastic characteristic K_F which are the mechanical characteristics of the coupling portion 101. The other configurations are the same as those of the longitudinal vibration damping unit 36 of the first embodiment.

FIG. 16 is a block diagram showing a configuration of the natural vibration damping unit 44 according to the third embodiment. As shown in FIG. 16, the natural vibration damping unit 44 of the present third embodiment includes a change rate calculation unit 301, a model response calculation unit 302, a difference calculation unit 303, a feature amount calculation unit 304, a dynamic characteristic setting unit 305, and a first vibration damping correction processing unit 306.

The change rate calculation unit 301 calculates a change rate of the motor torque T_m or a parameter corresponding to the change rate of the motor torque T_m. In this way, the change rate calculation unit 301 determines a traveling scene in which the longitudinal vibration is likely to occur.

In the present embodiment, the change rate calculation unit 301 calculates a change amount of the first torque target value T_m1* in a predetermined time (for example, one control cycle), that is, a time change rate of the first torque target value T_m1* (hereinafter, simply referred to as a change rate δT_m1* of the first torque target value T_m1*). More specifically, the change rate calculation unit 301 holds a previous value T_m1z* (not shown) of the first torque target value T_m1*. Therefore, the change rate calculation unit 301 calculates the change rate δT_m1* of the first torque target value T_m1* by comparing a current value of the first torque target value T_m1* and the previous value T_m1z*, or by calculating a difference between the current value of the first torque target value T_m1* and the previous value T_m1z*. The change rate δT_m1* of the first torque target value T_m1* is a parameter substantially representing the change rate (time change rate) of the motor torque T_m.

Then, the change rate calculation unit 301 determines a traveling scene in which the longitudinal vibration is likely to occur based on the change rate δT_m1* of the first torque target value T_m1*. Specifically, as the change rate δT_m1* of the first torque target value T_m1* increases, the longitudinal vibration is more likely to occur. In the case where the change rate δT_m1* of the first torque target value T_m1* is large, when there is an error in the elastic characteristic K_F or the viscosity characteristic C_F, or a modeling error, the longitudinal vibration is likely to remain even if the correction processing for reducing the longitudinal vibration in the first embodiment or the second embodiment is performed. Therefore, in the present embodiment, the change rate calculation unit 301 compares the change rate δT_m1* of the first torque target value T_m1* with a change rate threshold value δT_TH, and determines the traveling scene in which the longitudinal vibration is likely to occur when the change rate δT_m1* of the first torque target value T_m1* is larger than the change rate threshold value δT_TH. The change rate threshold value δT_TH is set in advance based on experiment, simulation, or the like with respect to the change rate δT_m1* of the first torque target value T_m1*.

The change rate calculation unit 301 sets a dynamic characteristic calculation flag FLG according to a comparison result between the change rate δT_m1* of the first torque target value T_m1* and the change rate threshold value δT_TH. The dynamic characteristic calculation flag FLG is an index indicating new calculation of the natural vibration frequency ω_t and the attenuation coefficient ζ_t (to identify the dynamic characteristic of the coupling portion 101) based on the longitudinal acceleration A_L1 and the like. The dynamic characteristic calculation flag FLG is set to, for example, “1” or “0”, and when the flag FLG is “1”, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are calculated based on the longitudinal acceleration A_L1 and the like. On the other hand, when the flag FLG is “0”, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are not calculated, and the previous value continues to be used. The change rate calculation unit 301 sets the dynamic characteristic calculation flag FLG to “1” when the change rate δ_Tm1* of the first torque target value T_m1* is larger than the change rate threshold value δT_TH. On the other hand, the change rate calculation unit 301 sets the dynamic characteristic calculation flag FLG to “0” when the change rate δ_Tm1* of the first torque target value T_m1* is equal to or less than the change rate threshold value δT_TH. The change rate calculation unit 301 inputs the dynamic characteristic calculation flag FLG to, for example, the model response calculation unit 302.

The model response calculation unit 302 calculates a model response A_L1-ref of the longitudinal acceleration A_L1 in the traveling scene in which at least the longitudinal vibration is likely to occur. In the present embodiment, the model response calculation unit 302 calculates the model response A_L1-ref of the longitudinal acceleration A_L1 when the dynamic characteristic calculation flag FLG is “1”.

The model response calculation unit 302 calculates the model response A_L1-ref of the longitudinal acceleration A_L1 based on the electric vehicle weight M₁, the coupled vehicle weight M₂, and the first torque target value T_m1*. Specifically, the model response calculation unit 302 calculates the model response A_L1-ref of the longitudinal acceleration A_L1 according to the following equation (25) by applying a model transmission characteristic G_pF-ref(s) expressed by the following equation (24) to the first torque target value T_m1*. The model response A_L1-ref of the longitudinal acceleration A_L1 is input to the difference calculation unit 303.

$\begin{array}{cc}\left[\mathrm{Math}.15\right]& \\ G_{pF-\mathrm{ref}}\left(s\right)=\frac{c_0}{\left(s+\alpha \right)·\left(s^2+2{\omega }_{p}s+{\omega }_p^2\right)}& \left(24\right)\end{array}$ $\begin{array}{cc}{A}_{L1-\mathrm{ref}}=\frac{1}{M}·G_{pF-\mathrm{ref}}\left(s\right)·T_{m1}^{*}& \left(25\right)\end{array}$

Note that the model transmission characteristic G_pF-ref(s) represents a model response from the motor torque T_m to the driving force F. The model transmission characteristic G_pF-ref(s) is obtained by setting ζ_p to 1 in the transmission characteristic G_pF(s) of the equation (17) as expressed by the equation (24). In the equation (25), “M” is a total weight of the electric vehicle 100 and the coupled vehicle 102 (=M₁+M₂). Except that the model transmission characteristic G_pF-ref(s) is used instead of the transmission characteristic G_pF(s), the model response A_L1-ref of the longitudinal acceleration A_L1 is calculated in the same manner as the longitudinal acceleration estimation value A_L1{circumflex over ( )}, and therefore the model response A_L1-ref of the longitudinal acceleration A_L1 is an estimation value. The above-described calculation method for the model response A_L1-ref of the longitudinal acceleration A_L1 is an example, and the model response A_L1-ref of the longitudinal acceleration A_L1 can be calculated using, for example, the vehicle model expressed by the above-described equation (19).

The difference calculation unit 303 calculates a difference ΔA_L1 between the longitudinal acceleration A_L1, which is an actual response, and the model response A_L1-ref. In the present embodiment, the difference calculation unit 303 calculates the difference ΔA_L1 by subtracting the model response A_L1-ref from the longitudinal acceleration A_L1, which is the actual response. After the dynamic characteristic calculation flag FLG becomes “1”, the difference calculation unit 303 continuously calculates the difference ΔA_L1 in at least a predetermined period PP. That is, the difference ΔA_L1 constitutes a data row that can change over time. The difference ΔA_L1 calculated by the difference calculation unit 303 is input to the feature amount calculation unit 304. Note that the “predetermined period PP” in which the difference ΔA_L1 is calculated is a time interval to the extent that time series data of the difference ΔA_L1 can be acquired to the extent that frequency analysis can be performed, and is determined in advance as, for example, from about 1 to several seconds by experiment, simulation, or the like. When a frequency range of the longitudinal vibration that may occur is identified by experiment, simulation, or the like, the difference calculation unit 303 can perform filtering processing of extracting components in the identified frequency range on the output difference ΔA_L1. In this case, noise superimposed on the difference ΔA_L1 (for example, noise corresponding to disturbance or the like) is removed, thereby improving accuracy of reducing the longitudinal vibration.

The feature amount calculation unit 304 identifies the dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled by calculating feature amounts of the longitudinal vibration generated by the coupled vehicle 102 being coupled to the coupling portion 101. In the present embodiment, the feature amount calculation unit 304 calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t, which are the feature amounts of the longitudinal vibration, based on the difference ΔA_L1 between the longitudinal acceleration A_L1, which is the actual response, and the model response A_L1-ref. That is, the feature amount calculation unit 304 holds the difference ΔA_L1 in the predetermined period PP, and calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the time series data. The feature amount calculation unit 304 can calculate the natural vibration frequency ω_t and the attenuation coefficient ζ_t by performing frequency analysis such as FFT (Fast Fourier Transform) on the time series data of the difference ΔA_L1. In the present embodiment, the feature amount calculation unit 304 simply calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the number of peaks (or bottoms) formed by the difference ΔA_L1 in the predetermined period PP and the change in amplitude thereof.

As described above, the feature amount calculation unit 304 does not use the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101 for calculating the natural vibration frequency ω_t and the attenuation coefficient ζ_t. That is, the feature amount calculation unit 304 identifies the dynamic characteristic of the coupling portion 101 by calculating the natural vibration frequency ω_t and the attenuation coefficient ζ_t which are feature amounts of the longitudinal vibration by a method different from that of the feature amount calculation unit 62 of the first embodiment. The natural vibration frequency ω_t and the attenuation coefficient ζ_t calculated by the feature amount calculation unit 304 are input to the dynamic characteristic setting unit 305.

Note that in the present embodiment, the feature amount calculation unit 304 compares the difference ΔA_L1 with an acceleration threshold value ΔA_TH. When the difference ΔA_L1 is larger than the acceleration threshold value ΔA_TH, the feature amount calculation unit 304 holds the time series data of the difference ΔA_L1 and calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the time series data of the difference ΔA_L1. That is, when the amplitude of the difference ΔA_L1 is entirely or partially larger than the amplitude determined by the acceleration threshold value ΔA_TH and a non-ignorable longitudinal vibration occurs, the feature amount calculation unit 304 calculates a new natural vibration frequency ω_t and a new attenuation coefficient ζ_t based on the time series data of the difference ΔA_L1. On the other hand, when the amplitude of the difference ΔA_L1 is equal to or less than the acceleration threshold value ΔA_TH in the entire range of the predetermined period PP, the feature amount calculation unit 304 determines that the scene is a scene in which longitudinal vibration, even if it occurs, can be substantially ignored. Then, the feature amount calculation unit 304 sets (resets) the dynamic characteristic calculation flag FLG to “0”, and does not calculate a new natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the difference ΔA_L1. The acceleration threshold value ΔA_TH is determined in advance by experiment, simulation, or the like.

The dynamic characteristic setting unit 305 sets the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be finally used as the dynamic characteristic of the coupling portion 101 by the first vibration damping correction processing unit 306.

In the present embodiment, an initial value ω_t0 (not shown) of the natural vibration frequency ω_t and an initial value ζ_t0 (not shown) of the attenuation coefficient ζ_t are determined in advance. Therefore, when the dynamic characteristic calculation flag FLG does not become “1” and the feature amount calculation unit 304 does not newly calculate the natural vibration frequency ω_t and the attenuation coefficient ζ_t, the dynamic characteristic setting unit 305 sets the initial values ω_t0 and ζ_t0 as the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be finally used as the dynamic characteristics of the coupling portion 101. The initial values ω_t0 and ζ_t0 are set based on, for example, the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101 and the vehicle model of the electric vehicle 100.

On the other hand, when the feature amount calculation unit 304 newly calculates the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the difference ΔA_L1 between the longitudinal acceleration A_L1, which is the actual response, and the model response A_L1-ref, the dynamic characteristic setting unit 305 compares the newly calculated natural vibration frequency ω_t and the attenuation coefficient ζ_t with existing constants. The existing constants are the natural vibration frequency ω_t and the attenuation coefficient ζ_t that are already used as the dynamic characteristics of the coupling portion 101, and are, for example, previous values ω_tz and ζ_tz (not shown). The previous values ω_tz and ζ_tz of the natural vibration frequency ω_t and the attenuation coefficient ζ_t are, for example, the initial values ω_t0 and ζ_t0.

More specifically, the dynamic characteristic setting unit 305 calculates a frequency deviation δω_t, which is a deviation between the natural vibration frequency ω_t newly calculated by the feature amount calculation unit 304 and the previous value ω_tz of the natural vibration frequency ω_t, and compares the calculated frequency deviation with a frequency threshold value δω_TH. The frequency threshold value δω_TH is determined in advance by experiment, simulation, or the like. When the frequency deviation δω_t is larger than the frequency threshold value δω_TH, the dynamic characteristic setting unit 305 updates the natural vibration frequency ω_t used as the dynamic characteristic of the coupling portion 101 to the natural vibration frequency ω_t newly calculated by the feature amount calculation unit 304. On the other hand, when the frequency deviation δω_t is equal to or less than the frequency threshold value δω_TH, the dynamic characteristic setting unit 305 holds the natural vibration frequency ω_t used as the dynamic characteristic of the coupling portion 101 at the previous value ω_tz. That is, the dynamic characteristic setting unit 305 updates the natural vibration frequency ω_t when the natural vibration frequency ω_t deviates from an existing constant by a predetermined degree or more.

Similarly, the dynamic characteristic setting unit 305 calculates an attenuation coefficient deviation δζ_t, which is a deviation between the attenuation coefficient ζ_t newly calculated by the feature amount calculation unit 304 and the previous value ζ_tz of the attenuation coefficient ζ_t, and compares the attenuation coefficient deviation δζ_t with the attenuation coefficient threshold value δζ_TH. The attenuation coefficient threshold value δζ_TH is determined in advance by experiment, simulation, or the like. When the attenuation coefficient deviation δζ_t is larger than the attenuation coefficient threshold value δζ_TH, the dynamic characteristic setting unit 305 updates the attenuation coefficient ζ_t used as the dynamic characteristic of the coupling portion 101 to the attenuation coefficient ζ_t newly calculated by the feature amount calculation unit 304. On the other hand, when the attenuation coefficient deviation δζ_t is equal to or less than the attenuation coefficient threshold value δζ_TH, the dynamic characteristic setting unit 305 holds the attenuation coefficient ζ_t used as the dynamic characteristic of the coupling portion 101 at the previous value ζ_tz. That is, the dynamic characteristic setting unit 305 updates the attenuation coefficient ζ_t when the attenuation coefficient ζ_t deviates from an existing constant by a predetermined degree or more. Note that the natural vibration frequency ω_t and the attenuation coefficient ζ_t can be updated or held independently of each other.

Note that the dynamic characteristic setting unit 305 limits the range of the natural vibration frequency ω_t that can be set. Specifically, the dynamic characteristic setting unit 305 defines an upper limit and a lower limit for the natural vibration frequency ω_t, and updates or holds the natural vibration frequency ω_t within a range between the upper limit and the lower limit. The upper limit and the lower limit of the natural vibration frequency ω_t determine a range of the natural vibration frequency ω_t that can be actually taken in towing traveling, and are determined in advance by experiment, simulation, or the like. Similarly, the dynamic characteristic setting unit 305 limits the range of the attenuation coefficient ζ_t that can be set. Specifically, the dynamic characteristic setting unit 305 determines an upper limit and a lower limit for the attenuation coefficient ζ_t, and updates or holds the attenuation coefficient ζ_t within a range between the upper limit and the lower limit. The upper limit and the lower limit of the attenuation coefficient ζ_t determine a range of the attenuation coefficient ζ_t that can be actually taken in towing traveling, and are determined in advance by experiment, simulation, or the like. In this way, the dynamic characteristic setting unit 305 limits the ranges of the natural vibration frequency ω_t and the attenuation coefficient ζ_t that can be set to realistic ranges. Accordingly, even if there is an estimation error in the natural vibration frequency ω_t and the attenuation coefficient ζ_t by the feature amount calculation unit 304, the behavior of the electric vehicle 100 does not become an unexpected unstable behavior due to the estimation error.

The first vibration damping correction processing unit 306 has the same configuration as the first vibration damping correction processing unit 63 (see FIG. 8) of the first embodiment except that the natural vibration frequency ω_t and the attenuation coefficient ζ_t are set by the dynamic characteristic setting unit 305. That is, the first vibration damping correction processing unit 306 calculates the second torque target value T_m2* by correcting the first torque target value T_m1* based on the natural vibration frequency ω_t and the attenuation coefficient ζ_t set by the dynamic characteristic setting unit 305.

Hereinafter, functions of setting the natural vibration frequency ω_t and the like in the electric vehicle 100 of the third embodiment configured as described above will be described.

FIG. 17 is a flowchart related to the updating of the natural vibration frequency and the like. As shown in FIG. 17, in step S301, the change rate calculation unit 301 calculates the change rate δT_m1* of the first torque target value T_m1*. In step S302, the change rate calculation unit 301 compares the change rate δT_m1* of the first torque target value T_m1* with the change rate threshold value δT_TH, and sets the dynamic characteristic calculation flag FLG.

In step S302, when the change rate δT_m1* of the first torque target value T_m1* is larger than the change rate threshold value δT_TH, the dynamic characteristic calculation flag FLG is set to “1”, and the process proceeds to step S303. This is a traveling scene in which there is a high possibility of occurrence of longitudinal vibration during towing traveling. On the other hand, in step S302, when the change rate δT_m1* of the first torque target value T_m1* is equal to or less than the change rate threshold value δT_TH, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are held at the previous values ω_tz and ζ_tz and the update processing shown in this flowchart is ended. This is a traveling scene in which the longitudinal vibration is unlikely to occur in towing traveling.

In step S303, the model response calculation unit 302 calculates the model response A_L1-ref of the longitudinal acceleration A_L1 based on the electric vehicle weight M₁, the coupled vehicle weight M₂, and the first torque target value T_m1*. In step S304, the difference calculation unit 303 calculates the difference ΔA_L1 between the longitudinal acceleration A_L1, which is an actual response, and the model response A_L1-ref. In step S305, the feature amount calculation unit 304 compares the difference ΔA_L1 with the acceleration threshold value ΔA_TH, and determines whether non-ignorable longitudinal vibration occurs.

In step S305, when the difference ΔA_L1 is equal to or less than the acceleration threshold value ΔA_TH and the generated longitudinal vibration is ignorable, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are held at the previous values ω_tz and ζ_tz and the update processing shown in this flowchart is ended.

On the other hand, in step S305, when the difference ΔA_L1 is larger than the acceleration threshold value ΔA_TH, the process proceeds to step S306, and the feature amount calculation unit 304 holds the difference ΔA_L1 in the predetermined period PP. Then, in step S307, the feature amount calculation unit 304 calculates a new natural vibration frequency ω_t and a new attenuation coefficient ζ_t based on the time series data of the held difference ΔA_L1.

Then, in step S308, the dynamic characteristic setting unit 305 calculates the frequency deviation δω_t, which is a deviation between the current value of the natural vibration frequency ω_t calculated by the feature amount calculation unit 304 and the previous value ω_tz of the natural vibration frequency ω_t, and compares the calculated frequency deviation with the frequency threshold value δω_TH. In step S308, when it is determined that the frequency deviation δω_t is larger than the frequency threshold value δω_TH and a non-ignorable error occurs in the natural vibration frequency ω_t, the process proceeds to step S309, and the dynamic characteristic setting unit 305 updates the natural vibration frequency ω_t of the coupling portion 101 to the new natural vibration frequency ω_t calculated by the feature amount calculation unit 304. On the other hand, in step S308, when it is determined that the frequency deviation δω_t is equal to or less than the frequency threshold value δω_TH and there is almost no error in the natural vibration frequency ω_t, step S309 is skipped.

Similarly, in step S310, the dynamic characteristic setting unit 305 calculates the attenuation coefficient deviation δζ_t, which is a deviation between the current value of the attenuation coefficient ζ_t calculated by the feature amount calculation unit 304 and the previous value ζ_tz of the attenuation coefficient ζ_t, and compares the attenuation coefficient deviation δζ_t with the attenuation coefficient threshold value δζ_TH. When it is determined in step S310 that the attenuation coefficient deviation δζ_t is larger than the attenuation coefficient threshold value δζ_TH and a non-ignorable error occurs in the attenuation coefficient ζ_t, the process proceeds to step S311, and the dynamic characteristic setting unit 305 updates the attenuation coefficient ζ_t of the coupling portion 101 to the new attenuation coefficient ζ_t calculated by the feature amount calculation unit 304. On the other hand, in step S310, when it is determined that the attenuation coefficient deviation δζt is equal to or less than the attenuation coefficient threshold value δζ_TH and there is almost no error in the attenuation coefficient ζ_t, step S311 is skipped. FIGS. 18(A)-18(H) are a time chart showing the longitudinal acceleration A_L1 and the like in a traveling scene in which the natural vibration frequency ω_t and the like are updated. FIG. 18(A), shows the first torque target value T_m1*. FIG. 18(B), shows the second torque target value T_m2*. FIG. 18(C), shows the drive shaft torque T_d. FIG. 18(D), shows the dynamic characteristic calculation flag FLG by a broken line, and shows the natural vibration frequency ω_t counted up by a solid line. FIG. 18(E), shows the motor rotation speed ω_m. FIG. 18(F), shows the longitudinal acceleration A_L1 (actual response) of the electric vehicle 100 by a solid line, and shows the model response A_L1-ref of the longitudinal acceleration A_L1 by a broken line. FIG. 18(G), for reference, shows the longitudinal acceleration A_L2 (actual response) of the coupled vehicle 102 by a solid line, and shows the model response A_L2-ref of the longitudinal acceleration A_L2 by a broken line. FIG. 18(H) shows the difference ΔA_L1 between the longitudinal acceleration A_L1 of the electric vehicle 100 and the model response A_L1-ref. Note that the horizontal axis of each graph in FIGS. 18(A)-18(H) represents time [s]. The time t₁ is a start time of an accelerator operation by the driver, and the time t₂ is the time after the predetermined period PP elapses from the time t₁.

As shown in FIG. 18(A), when the accelerator is operated at the time t₁, the first torque target value T_m1* is ramp input according to an operation amount. Accordingly, the dynamic characteristic calculation flag FLG is set to “1”. As shown in FIGS. 18(B) and 18(C), the second torque target value T_m2* and the drive shaft torque T_d increase in accordance with the first torque target value T_m1*. Note that as shown in FIG. 18(E), after the time t₁, the motor rotation speed ω_m generally increases with the elapse of time.

As shown in FIG. 18(F), in the present embodiment, vibration occurs in the longitudinal acceleration A_L1 of the electric vehicle 100. As shown in FIG. 18(G), vibration also occurs in the longitudinal acceleration A_L2 of the coupled vehicle 102 corresponding to the vibration of the longitudinal acceleration A_L1. That is, the longitudinal vibration occurs in the electric vehicle 100 and the coupled vehicle 102. Then, as shown by the difference ΔA_L1 in FIG. 18(H), the vibration of the longitudinal acceleration A_L1 has an amplitude which cannot be ignored with respect to the model response A_L1-ref, and therefore, as shown in FIG. 18(D), the dynamic characteristic calculation flag FLG is held at “1” from the time t₁ until the time t₂ after the predetermined period PP elapses. Therefore, the feature amount calculation unit 304 holds the time series data of the difference ΔA_L1 in the predetermined period PP, and counts the peaks of the time series data of the difference ΔA_L1 as shown in, for example, FIG. 18(D), thereby obtaining the natural vibration frequency ω_t of the longitudinal vibration. Similarly, the feature amount calculation unit 304 obtains the attenuation coefficient ζ_t of the longitudinal vibration based on a decrease in the amplitude of the difference ΔA_L1. Here, it is assumed that the natural vibration frequency ω_t and the attenuation coefficient ζ_t used as the dynamic characteristics of the coupling portion 101 are updated to the natural vibration frequency ω_t and the attenuation coefficient ζ_t obtained by the feature amount calculation unit 304.

FIGS. 19(A)-19(H) are a time chart showing the longitudinal acceleration A_L1 and the like after updating the natural vibration frequency ω_t and the like. FIG. 19(A), shows the first torque target value T_m1*. FIG. 19(B), shows the second torque target value T_m2*. FIG. 19(C), shows the drive shaft torque T_d. FIG. 19(D), shows the dynamic characteristic calculation flag FLG by a broken line, and shows the natural vibration frequency ω_t counted up by a solid line. FIG. 19(E), shows the motor rotation speed ω_m. FIG. 19(F), shows the longitudinal acceleration A_L1 (actual response) of the electric vehicle 100 by a solid line, and shows the model response A_L1-ref of the longitudinal acceleration A_L1 by a broken line. FIG. 19(G), for reference, shows the longitudinal acceleration A_L2 (actual response) of the coupled vehicle 102 by a solid line, and shows the model response A_L2-ref of the longitudinal acceleration A_L2 by a broken line. FIG. 19(H), shows the difference ΔA_L1 between the longitudinal acceleration A_L1 of the electric vehicle 100 and the model response A_L1-ref. Note that the horizontal axis of each graph in FIGS. 19(A)-19(H) represents time [s]. The time t₃ is a start time of an accelerator operation by the driver, and the time t₄ is the time after the predetermined period PP elapses from the time t₃.

As shown in FIG. 19(A), when the accelerator is operated at the time t₃ after the natural vibration frequency ω_t and the like are updated, the first torque target value T_m1* is ramp input according to an operation amount. Accordingly, the dynamic characteristic calculation flag FLG is set to “1”. As shown in FIGS. 19(B) and 19(C), the second torque target value T_m2* and the drive shaft torque T_d increase in accordance with the first torque target value T_m1*. Note that as shown in FIG. 19(E), after the time t₃, the motor rotation speed ω_m generally increases with the elapse of time. These behaviors are similar to those before the natural vibration frequency ω_t and the like are updated (FIGS. 18(A)-18(H)).

On the other hand, as shown in FIG. 19(F), since the natural vibration frequency ω_t and the like are updated, the vibration of the longitudinal acceleration A_L1 of the electric vehicle 100 is reduced. As shown in FIG. 19(G), the vibration of the longitudinal acceleration A_L2 of the coupled vehicle 102 is also reduced. Therefore, as shown in FIG. 19(H), the difference ΔA_L1 does not have a significant amplitude. Therefore, the dynamic characteristic calculation flag FLG is reset by the feature amount calculation unit 304, and is set to “0” at least from the time t₃ to the time t₄. Accordingly, the feature amount calculation unit 304 does not calculate a new natural vibration frequency ω_t and a new attenuation coefficient ζ_t based on the difference ΔA_L1, and the natural vibration frequency ω_t and the attenuation coefficient ζ_t used as the dynamic characteristics of the coupling portion 101 are held at the previous values ω_tz and ζ_tz. As a result, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is appropriately reduced even later.

As described above, in the electric vehicle 100 according to the third embodiment, the natural vibration frequency ω_t and the attenuation coefficient ζ_t which are dynamic characteristics of the coupling portion 101 are calculated based on the difference ΔA_L1 between the longitudinal acceleration A_L1 which is an actual response and the model response A_L1-ref. Then, the first torque target value T_m1* is corrected based on the natural vibration frequency ω_t and the attenuation coefficient ζ_t calculated based on this difference ΔA_L1. Accordingly, in the electric vehicle 100 according to the third embodiment, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is appropriately reduced even when there is an error in the known values of the mechanical characteristics (the viscosity characteristic C_F, the elastic characteristic K_F, and the like) of the coupling portion 101 or there is a modeling error in the vehicle model.

Note that in the control according to the third embodiment, as in the first embodiment, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are directly calculated, and the compensation for the longitudinal vibration unique to towing traveling and the compensation for the power transmission mechanism vibration are separately performed. However, the control according to the third embodiment can be performed in a configuration in which the compensation for the longitudinal vibration unique to towing traveling and the compensation for the power transmission mechanism vibration are substantially integrally performed, as in the second embodiment.

FIG. 20 is a block diagram showing a partial configuration of the vibration damping control unit 33 in a case where the compensation for the longitudinal vibration unique to towing traveling and the compensation for the power transmission mechanism vibration are substantially integrally performed. When the compensation for the longitudinal vibration unique to towing traveling and the compensation for the power transmission mechanism vibration are substantially integrally performed, as shown in FIG. 20, the feedforward compensation unit 64 of the vibration damping control unit 33 is configured to directly calculate the fourth torque target value T_m4* based on the first torque target value T_m1*. Specifically, the feedforward compensation unit 64 includes the electric vehicle model 91, the coupled vehicle model 92, the compensation torque calculation unit 93, and the vibration damping correction processing unit 94. That is, the overall configuration is the same as that of the second embodiment.

When the control of the third embodiment is performed in the vibration damping control unit 33 (feedforward compensation unit 64) in this way, as shown in FIG. 20, the viscosity characteristic C_F and the elastic characteristic K_F, the electric vehicle weight M₁, or the coupled vehicle weight M₂ are variable according to the natural vibration frequency ω_t and the attenuation coefficient ζ_t set by the dynamic characteristic setting unit 305. That is, the dynamic characteristic setting unit 305 is configured to correct the viscosity characteristic C_F and the elastic characteristic K_F included in the coupled vehicle model 92, the electric vehicle weight M₁ included in the electric vehicle model 91, and the coupled vehicle weight M₂ included in the coupled vehicle model 92, according to the set natural vibration frequency ω_t and the set attenuation coefficient ζ_t. The dynamic characteristic setting unit 305 can inversely calculate the viscosity characteristic C_F and the elastic characteristic K_F corresponding to the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be set, the electric vehicle weight M₁, or the coupled vehicle weight M₂ in accordance with the equations of motion or the vehicle model described above.

For example, when the natural vibration frequency ω_t is updated, the dynamic characteristic setting unit 305 calculates the elastic characteristic K_F (hereinafter, referred to as a corrected elastic characteristic K_F′ (not shown)) according to the natural vibration frequency ω_t to be set. Then, the dynamic characteristic setting unit 305 updates the elastic characteristic K_F of the coupled vehicle model 92 to this corrected elastic characteristic K_F′. Accordingly, similarly to the control of the third embodiment, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is appropriately reduced even when there is an error in the known values of the mechanical characteristics (the viscosity characteristic C_F, the elastic characteristic K_F, and the like) of the coupling portion 101 or there is a modeling error in the vehicle model.

For example, when the attenuation coefficient ζ_t is updated, the dynamic characteristic setting unit 305 calculates the viscosity characteristic C_F (hereinafter, referred to as a corrected viscosity characteristic C_F′ (not shown)) corresponding to the attenuation coefficient ζ_t to be set. Then, the dynamic characteristic setting unit 305 updates the viscosity characteristic C_F of the coupled vehicle model 92 to the corrected viscosity characteristic C_F′. Accordingly, similarly to the control of the third embodiment, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is appropriately reduced even when there is an error in the known values of the mechanical characteristics (the viscosity characteristic C_F, the elastic characteristic K_F, and the like) of the coupling portion 101 or there is a modeling error in the vehicle model.

Since the natural vibration frequency ω_t is a parameter directly related to the elastic characteristic K_F, as described above, when updating the natural vibration frequency ω_t, it is particularly preferable that the dynamic characteristic setting unit 305 updates the elastic characteristic K_F to the corrected elastic characteristic K_F′. Since the attenuation coefficient ζ_t is a parameter directly related to the viscosity characteristic C_F, it is particularly preferable that the dynamic characteristic setting unit 305 updates the viscosity characteristic C_F to the corrected viscosity characteristic C_F′ when update the attenuation coefficient ζ_t as described above. However, the relative vehicle speed Δv₁₂ reflecting the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be set may be obtained in the vibration damping control unit 33 that substantially integrally performs the compensation for the longitudinal vibration unique to towing traveling and the compensation for the power transmission mechanism vibration. Therefore, instead of directly updating the elastic characteristic K_F and the viscosity characteristic C_F, other parameters may be updated according to the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be set. For example, the dynamic characteristic setting unit 305 can calculate the electric vehicle weight M₁ (hereinafter, referred to as a corrected electric vehicle weight M₁′ (not shown)) corresponding to the natural vibration frequency ω_t or the attenuation coefficient ζ_t to be set, and update the electric vehicle weight M₁ included in the electric vehicle model 91 to the corrected electric vehicle weight M₁′. Similarly, the dynamic characteristic setting unit 305 can calculate the coupled vehicle weight M₂ (hereinafter, referred to as a corrected coupled vehicle weight M₂′ (not shown)) corresponding to the natural vibration frequency ω_t or the attenuation coefficient ζ_t to be set, and update the coupled vehicle weight M₂ included in the coupled vehicle model 92 to the corrected coupled vehicle weight M₂′. In these cases, similarly to the control of the third embodiment, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is appropriately reduced even when there is an error in the known values of the mechanical characteristics (the viscosity characteristic C_F, the elastic characteristic K_F, and the like) of the coupling portion 101 or there is a modeling error in the vehicle model.

Note that the dynamic characteristic setting unit 305 can calculate two or more parameters among the corrected elastic characteristic K_F′, the corrected viscosity characteristic C_F′, the corrected electric vehicle weight M₁′, and the corrected coupled vehicle weight M₂′ according to the natural vibration frequency ω_t or the attenuation coefficient ζ_t to be set, and update the corresponding plurality of parameters. However, the dynamic characteristic setting unit 305 preferably corrects the elastic characteristic K_F based on at least the natural vibration frequency ω_t to be set. The dynamic characteristic setting unit 305 preferably corrects the elastic characteristic K_F and the viscosity characteristic C_F based on the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be set.

As described above, the third embodiment can be implemented independently instead of the first embodiment or the second embodiment. However, the third embodiment can be implemented together with the first embodiment or the second embodiment. When the control according to the third embodiment is performed together with the control according to the first embodiment, for example, the initial values ω_t0 and ζ_t0 of the natural vibration frequency ω_t and the attenuation coefficient ζ_t according to the third embodiment may be changed to the natural vibration frequency ω_t and the attenuation coefficient t calculated by the feature amount calculation unit 62 according to the first embodiment. When the control according to the third embodiment is performed together with the control according to the second embodiment, for example, as in the second embodiment, the electric vehicle weight M₁ and the coupled vehicle weight M₂ may be set according to respective estimation results thereof, and the elastic characteristic K_F, the viscosity characteristic C_F, the electric vehicle weight M₁, and the coupled vehicle weight M₂ may be updated according to the natural vibration frequency ω_t and the attenuation coefficient ζ_t to be set.

As described above, when the control for the electric vehicle 100 according to the third embodiment is performed together with the first embodiment or the second embodiment, the state in which the longitudinal vibration is reduced by the control of the first embodiment or the second embodiment is a standard state. The control according to the third embodiment is configured to particularly accurately reduce (compensate for) the errors included in the known values of the elastic characteristic K_F and the viscosity characteristic C_F, or the longitudinal vibration still remaining due to the modeling error of the electric vehicle 100 or the like. Accordingly, when the control for the electric vehicle 100 according to the third embodiment is performed together with the control according to the first embodiment or the second embodiment, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 can be particularly suitably reduced.

Modifications

Note that in each of the above embodiments, the electric vehicle 100 includes the towing traveling switch 28 as described above, and it is determined whether the coupled vehicle 102 is coupled to the coupling portion 101 based on the towing traveling signal SW_T output from the towing traveling switch 28, but the present invention is not limited thereto. For example, the towing traveling switch 28 may be provided in the coupling portion 101. In this case, when the coupled vehicle 102 is coupled to the coupling portion 101, the towing traveling signal SW_T automatically becomes ON without any operation by the driver or the like. When the electric vehicle 100 is equipped with a rear camera that captures an image of a rear side of the vehicle, the coupled vehicle 102 may be recognized using the image captured by the rear camera, and the motor controller 13 may determine whether the coupled vehicle 102 is coupled to the coupling portion 101 based on the recognition result. When the coupled vehicle weight M₂ becomes a significant value, specifically, for example, when the estimated weight M₂ becomes a predetermined threshold value or more, the motor controller 13 may determine that the coupled vehicle 102 is coupled to the coupling portion 101. In this case, whether the coupled vehicle 102 is coupled to the coupling portion 101 is simply and directly determined based on whether the coupled vehicle weight M₂ is detected, regardless of the operation of the towing traveling switch 28 by the driver or the like, the recognition processing of the coupled vehicle 102 by the rear camera, or the like.

In each of the above embodiments, the weights M₁ and M₂ of the electric vehicle 100 and the coupled vehicle 102 are estimated, but these values can also be set as default values. For example, when a change in the weight M₁ of the electric vehicle 100 is considered to be ignorable, the design weight M_ini of the electric vehicle 100 can be used. The weight M₂ of the coupled vehicle 102 can be a selectable manual input or the like according to the vehicle type or the like. As described above, even in the case where the weights M₁ and M₂ of the electric vehicle 100 and the coupled vehicle 102 are not sequentially estimated and are set to default values, the vibration damping effect against the longitudinal vibration unique to the towing traveling can be obtained as in the above embodiments. However, as in each of the above embodiments, by sequentially estimating the weights M₁ and M₂ of the electric vehicle 100 and the coupled vehicle 102, the longitudinal vibration unique to the towing traveling is particularly accurately reduced.

In each of the above embodiments, the power transmission mechanism vibration is reduced, but when the electric vehicle 100 does not include the drive shaft 22, the damping control of the power transmission mechanism vibration can be omitted. That is, the power transmission mechanism vibration damping unit 37 may be omitted.

In addition, in each of the above embodiments, in order to identify the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled, the viscosity characteristics C_F and the elastic characteristics K_F are considered as the mechanical characteristics of the coupling portion 101, but the present invention is not limited thereto. When the coupling portion 101 includes a gear, backlash of the gear can be considered. Accordingly, the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled are particularly accurately specified. However, by considering at least the viscosity characteristic C_F and the elastic characteristic K_F as the mechanical characteristics of the coupling portion 101, the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled are identified to such an extent that the longitudinal vibration unique to towing traveling can be easily and sufficiently reduced.

In each of the above embodiments, the dynamic characteristics of the coupling portion 101 are identified by the viscosity characteristic C_F and the elastic characteristic K_F, but the dynamic characteristic of the coupling portion 101 may be identified by either the elastic characteristic K_F or the viscosity characteristic C_F. However, in order to particularly accurately reduce the longitudinal vibration, it is preferable that the dynamic characteristic of the coupling portion 101 is identified by at least the elastic characteristic K_F. That is, in the first vibration damping correction processing, the component of the natural vibration frequency ω_t included in the first torque target value T_m1* may be reduced based on at least the natural vibration frequency ω_t. In the first vibration damping correction processing, the component of the natural vibration frequency ω_t included in the first torque target value T_m1* may also be reduced according to the attenuation coefficient ζ_t based on the natural vibration frequency ω_t and the attenuation coefficient ζ_t.

In each of the above embodiments, the electric vehicle 100 calculates the first torque target value T_m1* based on the vehicle operation (accelerator operation or the like) by the driver, but the present invention is not limited thereto. The electric vehicle 100 may perform a vehicle operation. For example, in a case where the electric vehicle 100 is an autonomous driving vehicle or in a case where the electric vehicle 100 assists the vehicle operation by the driver as necessary, the electric vehicle 100 can calculate (determine) the first torque target value T_m1* according to the determination of the electric vehicle 100 itself without depending on the vehicle operation by the driver. That is, the “vehicle operation” includes, in addition to the accelerator operation or the like by the driver, an operation of setting or changing a vehicle variable or the like that can be set or changed by an operation by the driver according to the determination of the electric vehicle 100 itself.

In each of the above embodiments, since the longitudinal vibration of the electric vehicle 100 and the longitudinal vibration of the coupled vehicle 102 are relative to each other and are related to each other, calculation using the longitudinal acceleration A_L1 of the electric vehicle 100 or the like can be replaced with calculation using the longitudinal acceleration A_L2 of the coupled vehicle 102 or the like. In each of the above embodiments, calculation using the longitudinal acceleration A_L1 of the electric vehicle 100 or the like can be replaced with calculation using the longitudinal acceleration A_L1 of the electric vehicle 100 and the longitudinal acceleration A_L2 of the coupled vehicle 102 or the like. However, depending on the specific coupled vehicle 102, the longitudinal acceleration A_L2 may not necessarily be acquired, and it is preferable that the longitudinal acceleration A_L1 of the electric vehicle 100 is acquired and used as in the above embodiments.

As described above, the control method for an electric vehicle according to each of the above embodiments and modification is a control method for the electric vehicle 100 that includes the motor 10 as a drive source and the coupling portion 101 coupled to the other vehicle, and travels while towing a coupled vehicle 102, which is the other vehicle coupled to the coupling portion 101. In this control method, the basic torque target value (first torque target value T_m1*) representing the torque (motor torque T_m) to be output by the motor 10 is calculated based on the vehicle operation (accelerator opening A_po or the like). The final torque command value (sixth torque target value T_m6*), which is a final command value for the motor torque T_m, is calculated by performing correction processing (first vibration damping correction processing) for reducing the longitudinal vibration component generated in the electric vehicle 100 due to the coupled vehicle 102 being coupled to the coupling portion 101 on the basic torque target value, based on a dynamic characteristic of the coupling portion 101 to which the coupled vehicle 102 is coupled. Then, the motor 10 is controlled based on this final torque command value.

In this way, by performing the correction processing for reducing the longitudinal vibration unique to the towing traveling, the rise of the torque required by the vehicle operation and the smooth acceleration are achieved. Especially, although it is difficult to achieve both the rise of the torque as required and the smooth acceleration when the longitudinal vibration is generated during towing traveling, the rise of the torque as required and the smooth acceleration can be achieved when the longitudinal vibration is reduced by the first vibration damping correction processing.

In the control method for an electric vehicle according to each of the above embodiments and modification, in the correction processing (first vibration damping correction processing) for reducing the longitudinal vibration component, the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled are identified based on the weight of the electric vehicle 100 (electric vehicle weight M₁), the weight of the coupled vehicle 102 (coupled vehicle weight M₂), and the mechanical characteristics of the coupling portion 101 (viscosity characteristic C_F and elastic characteristic K_F). In this way, the longitudinal vibration unique to the towing traveling is particularly accurately reduced by identifying the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled based on the electric vehicle weight M₁, the coupled vehicle weight M₂, and the mechanical characteristics of the coupling portion 101.

In the control method for an electric vehicle according to each of the above embodiments and modification, the mechanical characteristics of the coupling portion 101 include at least the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101. That is, in order to identify the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled, at least the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101 are considered. Therefore, by using at least the viscosity characteristic C_F and the elastic characteristic K_F of the coupling portion 101, the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled are identified easily and accurately.

In the control method for an electric vehicle according to each of the above embodiments and modification, specifically, the total weight M{circumflex over ( )} of the electric vehicle 100 and the coupled vehicle 102 is estimated based on the longitudinal acceleration (longitudinal acceleration A_L1) of the electric vehicle 100. The weight of the electric vehicle 100 (electric vehicle weight M₁) is estimated based on the stroke amounts ST_FL, ST_FR, ST_RL, and ST_RR of the suspensions included in the electric vehicle 100. Then, the weight of the coupled vehicle 102 (coupled vehicle weight M₂) is estimated by subtracting the weight of the electric vehicle 100 (electric vehicle weight M₁) from the total weight M{circumflex over ( )}.

The coupled vehicle weight M₂ varies depending on the specific coupled vehicle 102. Accordingly, as described above, by sequentially estimating the electric vehicle weight M₁ and the coupled vehicle weight M₂ during towing traveling, the longitudinal vibration unique to the towing traveling is accurately reduced regardless of the specific coupled vehicle 102.

In the control method for an electric vehicle according to each of the above embodiments and modification, it is determined whether the coupled vehicle 102 is coupled to the coupling portion 101. When it is determined that the coupled vehicle 102 is not coupled to the coupling portion 101, the correction processing for reducing the longitudinal vibration unique to the towing traveling is not executed. On the other hand, when it is determined that the coupled vehicle 102 is coupled to the coupling portion 101, the correction processing for reducing the longitudinal vibration unique to the towing traveling is executed. In this way, the control stability is improved by determining whether the vehicle is in the towing traveling scene and performing the correction processing for reducing the longitudinal vibration unique thereto as necessary. The load of calculation and the like is reduced.

In the control method for an electric vehicle according to the first embodiment and the modification, in the correction processing (first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling, the natural vibration frequency ω_t of the longitudinal vibration component is calculated based on the weight of the electric vehicle 100 (electric vehicle weight M₁), the weight of the coupled vehicle 102 (coupled vehicle weight M₂), and the mechanical characteristics of the coupling portion 101 (viscosity characteristic C_F and elastic characteristic K_F). Then, the component of the natural vibration frequency ω_t included in the basic torque target value (first torque target value T_m1*) is reduced. As described above, by identifying the natural vibration frequency ω_t, the longitudinal vibration unique to towing traveling is particularly accurately reduced.

In the control method for an electric vehicle according to the first embodiment and the modification, in the correction processing (first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling, the attenuation coefficient ζ_t of the longitudinal vibration component is further calculated based on the weight of the electric vehicle 100 (electric vehicle weight M₁), the weight of the coupled vehicle 102 (coupled vehicle weight M₂), and the mechanical characteristics of the coupling portion 101 (viscosity characteristic C_F and elastic characteristic K_F). Then, the component of the natural vibration frequency ω_t included in the basic torque target value (first torque target value T_m1*) is reduced according to the attenuation coefficient ζ_t. In this way, by reducing the component of the natural vibration frequency ω_t according to the attenuation coefficient ζ_t, the longitudinal vibration unique to the towing traveling is particularly efficiently reduced.

In the control method for an electric vehicle according to the first embodiment and the modification, the correction processing (first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is particularly configured so that the dynamic characteristics of the coupling portion 101 are identified and the component of the natural vibration frequency ω_t included in the basic torque target value (first torque target value T_m1*) is reduced according to the attenuation coefficient ζ_t, by calculating the natural vibration frequency ω_t and the attenuation coefficient ζ_t of the longitudinal vibration component based on the weight of the electric vehicle 100 (electric vehicle weight M₁), the weight of the coupled vehicle 102 (coupled vehicle weight M₂), and the mechanical characteristics of the coupling portion 101 (viscosity characteristic C_F and elastic characteristic K_F). In this way, the dynamic characteristics of the coupling portion 101 are identified by the natural vibration frequency ω_t and the attenuation coefficient ζ_t, and the component of the natural vibration frequency or is reduced according to the attenuation coefficient ζ_t, and thus the longitudinal vibration unique to the towing traveling is particularly suitably reduced.

In the control method for an electric vehicle according to the first embodiment and the modification, the correction processing (first vibration damping correction processing) for reducing the longitudinal vibration unique to towing traveling is performed by processing the basic torque target value (first torque target value T_m1*) using a band-stop filter in which the natural vibration frequency or is set as a center frequency for reduction and the attenuation coefficient (ζ in the equation (20)) is set to 1 or more. In this way, by performing the correction processing for reducing the longitudinal vibration unique to the towing traveling by the band-stop filter, the longitudinal vibration unique to the towing traveling is reduced, and a high-speed response is achieved.

In the control method for an electric vehicle according to the second embodiment and the modification, in the correction processing (vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling, the relative vehicle speed Δv₁₂ between the electric vehicle 100 and the coupled vehicle 102 is calculated based on the dynamic characteristics of the coupling portion 101 to which the coupled vehicle 102 is coupled. Based on the relative vehicle speed Δv₁₂, a correction torque (second compensation torque T_FF2*) corresponding to the longitudinal vibration component is calculated. Then, the basic torque target value (first torque target value T_m1*) is corrected using the correction torque (second compensation torque T_FF2*). In this way, when the basic torque target value (first torque target value T_m1*) is corrected using the correction torque (second compensation torque T_FF2*) based on the relative vehicle speed Δv₁₂, the longitudinal vibration unique to the towing traveling is reduced, and particularly, a high-speed response is achieved.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to identify the dynamic characteristics of the coupling portion 101 by calculating the natural vibration frequency ω_t of the longitudinal vibration component, or the attenuation coefficient ζ_t, or both the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the longitudinal acceleration (A_L1) of the electric vehicle 100. Further, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) is configured to reduce the component of the natural vibration frequency ω_t included in the basic torque target value (the first torque target value T_m1*). In this way, in a case where the natural vibration frequency ω_t and/or the attenuation coefficient ζ_t is identified based on the longitudinal acceleration A_L1 of the electric vehicle 100 and the component of the natural vibration frequency ω_t is reduced, the longitudinal vibration unique to the towing traveling is accurately reduced even when an error is included in the mechanical characteristics (the viscosity characteristic C_F and the elastic characteristic K_F) of the coupling portion 101 or the vehicle model of the electric vehicle 100. Since the electric vehicle 100 normally includes a sensor that detects the longitudinal acceleration A_L1, according to the control method for an electric vehicle according to the third embodiment and the modification, it is not necessary to additionally provide a new sensor or the like in order to reduce the longitudinal vibration unique to towing traveling, and the longitudinal vibration is accurately reduced without increasing a cost.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is particularly configured to identify the dynamic characteristics of the coupling portion 101 and reduce the component of the natural vibration frequency ω_t included in the basic torque target value (first torque target value T_m1*) according to the attenuation coefficient ζ_t by calculating the natural vibration frequency ω_t of the longitudinal vibration component and the attenuation coefficient ζ_t based on the longitudinal acceleration (longitudinal acceleration A_L1) of the electric vehicle 100. In this way, in a case where the natural vibration frequency ω_t and the attenuation coefficient ζ_t are identified based on the longitudinal acceleration (A_L1) of the electric vehicle 100 and the component of the natural vibration frequency ω_t is reduced according to the attenuation coefficient ζ_t, the longitudinal vibration unique to the towing traveling is particularly accurately reduced even when an error is included in the mechanical characteristics (C_F and K_F) of the coupling portion 101 or the vehicle model of the electric vehicle 100.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to be executed when the change rate (δT_m1*) of the basic torque target value (the first torque target value T_m1*) is greater than the predetermined threshold value (δT_TH). In this way, the traveling scene in which the longitudinal vibration is likely to occur is determined based on the change rate δT_m1* of the first torque target value T_m1*, and at least the first vibration damping correction processing is executed in such a traveling scene, and therefore the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is particularly accurately reduced.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to calculate the model response A_L1-ref of the longitudinal acceleration (A_L1), and to calculate the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the difference ΔA_L11 between the acquired longitudinal acceleration (A_L1) and the model response A_L1-ref of the longitudinal acceleration (A_L1). In this way, by calculating the natural vibration frequency ω_t and the attenuation coefficient ζ_t based on the difference ΔA_L1, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are estimated accurately in accordance with the actual condition regardless of known values of the elastic characteristic K_F and the viscosity characteristic C_F, a modeling error, or the like. As a result, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is particularly appropriately reduced. Note that when the control for an electric vehicle according to the third embodiment and the modification are performed together with the control of the first embodiment or the second embodiment, the remaining longitudinal vibration can be accurately reduced even when the control of the first embodiment or the second embodiment are performed.

In the control method for an electric vehicle according to the third embodiment and the modification, the model response A_L1-ref of the longitudinal acceleration (A_L1) is calculated based on the basic torque target value (first torque target value T_m1*), the weight of the electric vehicle 100 (M₁), and the weight of the coupled vehicle 102 (M₂). In this way, the model response A_L1-ref of the longitudinal acceleration A_L1 is calculated based on the first torque target value T_m1* and the estimation values of the electric vehicle weight M₁ and the coupled vehicle weight M₂, so that the model response A_L1-ref of the longitudinal acceleration A_L1 in accordance with the actual condition can be acquired. Therefore, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is particularly appropriately reduced.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to calculate the natural vibration frequency ω_t and the attenuation coefficient ζ_t by performing the frequency analysis processing on the difference ΔA_L1 between the acquired longitudinal acceleration (A_L1) and the model response A_L1-ref of the longitudinal acceleration (A_L1). In this way, the natural vibration frequency ω_t and the attenuation coefficient ζ_t are calculated by performing the frequency analysis processing on the difference ΔA_L1 between the longitudinal acceleration A_L1, which is an actual response, and the model response A_L1-ref of the longitudinal acceleration A_L1, so that particularly accurate natural vibration frequency ω_t and attenuation coefficient ζ_t corresponding to actually occurring longitudinal vibration can be obtained. Therefore, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is particularly appropriately reduced.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to update the dynamic characteristics of the coupling portion 101 (ω_t, ζ_t) when the dynamic characteristics change more than the predetermined threshold values (δω_TH, δζ_TH) with respect to the previous value (ω_tz, ζ_tz). In this way, by updating the natural vibration frequency ω_t and the attenuation coefficient ζ_t when the natural vibration frequency ω_t and the attenuation coefficient ζ_t deviate from the existing constants to predetermined degrees or more, a severe change in the natural vibration frequency ω_t and the attenuation coefficient ζ_t is prevented, and the natural vibration frequency ω_t and the attenuation coefficient ζ_t are appropriately updated when necessary. Therefore, the longitudinal vibration of the electric vehicle 100 and the coupled vehicle 102 is particularly appropriately reduced.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to limit the ranges of the natural vibration frequency ω_t and the attenuation coefficient ζ_t that can be set to predetermined ranges (ranges of values that can be actually taken during the towing traveling). In this way, by limiting the ranges of the natural vibration frequency ω_t and the attenuation coefficient ζ_t to realistic ranges, even when an estimation error is included in the natural vibration frequency ω_t or the attenuation coefficient ζ_t, the behavior of the electric vehicle 100 is prevented from being unstable due to the estimation error.

In the control method for an electric vehicle according to the third embodiment and the modification, the correction processing (first vibration damping correction processing, or vibration damping correction processing including the first vibration damping correction processing) for reducing the longitudinal vibration unique to the towing traveling is configured to calculate the natural vibration frequency ω_t and attenuation coefficient ζ_t of the longitudinal vibration component based on the longitudinal acceleration (A_L1) of the electric vehicle 100, correct the mechanical characteristics (K_F, C_F) of the coupling portion 101 or the weight (M₁, M₂) of the electric vehicle 100 or the coupled vehicle 102 based on the natural vibration frequency ω_t and attenuation coefficient ζ_t, calculate the relative vehicle speed Δv₁₂ between the electric vehicle 100 and the coupled vehicle 102 based on the corrected mechanical characteristics (K_F′, C_F′) of the coupling portion 101 or the corrected weight (M₁′, M₂′) of the electric vehicle 100 or the coupled vehicle 102, calculate the correction torque (T_FF2*) for the longitudinal vibration component based on the relative vehicle speed Δv₁₂, and correct the basic torque target value (first torque target value T_m1*) using the correction torque (T_FF2*). In this way, in a case where the natural vibration frequency ω_t and/or the attenuation coefficient t is identified based on the longitudinal acceleration (A_L1) of the electric vehicle 100 and the component of the natural vibration frequency ω_t is reduced, the longitudinal vibration unique to the towing traveling is accurately reduced even when an error is included in the mechanical characteristics (the viscosity characteristic C_F and the elastic characteristic K_F) of the coupling portion 101 or the vehicle model of the electric vehicle 100. Since the electric vehicle 100 normally includes a sensor that detects the longitudinal acceleration A_L1, according to the control method for an electric vehicle according to the third embodiment and the modification, it is not necessary to additionally provide a new sensor or the like in order to reduce the longitudinal vibration unique to towing traveling, and the longitudinal vibration is accurately reduced without increasing a cost.

Although the embodiments and modification of the present invention have been described above, configurations described in the above embodiments and the modification are merely a part of application examples of the present invention, and are not intended to limit the technical scope of the present invention.

Claims

1. A control method for an electric vehicle that includes a motor as a drive source and a coupling portion coupled to another vehicle, and travels while towing a coupled vehicle, which is the other vehicle coupled to the coupling portion, the control method comprising: calculating a basic torque target value representing a torque to be output by the motor based on a vehicle operation;calculating a final torque command value, which is a final command value for the torque, by performing correction processing for reducing a longitudinal vibration component generated in the electric vehicle due to the coupled vehicle being coupled to the coupling portion on the basic torque target value, based on a dynamic characteristic of the coupling portion to which the coupled vehicle is coupled; andcontrolling the motor based on the final torque command value.

2. The control method for an electric vehicle according to claim 1, wherein the correction processing includes identifying the dynamic characteristic based on a weight of the electric vehicle, a weight of the coupled vehicle, and mechanical characteristics of the coupling portion.

3. The control method for an electric vehicle according to claim 2, wherein the mechanical characteristics of the coupling portion include at least a viscosity characteristic and an elastic characteristic of the coupling portion.

4. The control method for an electric vehicle according to claim 2, further comprising: estimating a total weight of the electric vehicle and the coupled vehicle based on a longitudinal acceleration of the electric vehicle;estimating the weight of the electric vehicle based on a stroke amount of a suspension included in the electric vehicle; andestimating the weight of the coupled vehicle by subtracting the weight of the electric vehicle from the total weight.

5. The control method for an electric vehicle according to claim 1, wherein the correction processing includes: identifying the dynamic characteristic by calculating a natural vibration frequency and an attenuation coefficient of the longitudinal vibration component based on a weight of the electric vehicle, a weight of the coupled vehicle, and mechanical characteristics of the coupling portion; andreducing a component of the natural vibration frequency included in the basic torque target value according to the attenuation coefficient.

6. The control method for an electric vehicle according to claim 5, wherein the correction processing is performed by processing the basic torque target value using a band-stop filter in which the natural vibration frequency is set as a center frequency for reduction and the attenuation coefficient is set to 1 or more.

7. The control method for an electric vehicle according to claim 1, wherein the correction processing includes: calculating a relative vehicle speed between the electric vehicle and the coupled vehicle based on the dynamic characteristic;calculating a correction torque for the longitudinal vibration component based on the relative vehicle speed; and correcting the basic torque target value using the correction torque.

8. The control method for an electric vehicle according to claim 1, wherein the correction processing includes: identifying the dynamic characteristic by calculating a natural vibration frequency and an attenuation coefficient of the longitudinal vibration component based on a longitudinal acceleration of the electric vehicle; andreducing a component of the natural vibration frequency included in the basic torque target value according to the attenuation coefficient.

9. The control method for an electric vehicle according to claim 8, wherein the correction processing is executed at least when a change rate of the basic torque target value is greater than a predetermined threshold value.

10. The control method for an electric vehicle according to claim 8, wherein the correction processing includes: calculating a model response of the longitudinal acceleration; andcalculating the natural vibration frequency and the attenuation coefficient based on a difference between an acquired longitudinal acceleration and the model response of the longitudinal acceleration.

11. The control method for an electric vehicle according to claim 10, wherein the model response of the longitudinal acceleration is calculated based on the basic torque target value, a weight of the electric vehicle, and a weight of the coupled vehicle.

12. The control method for an electric vehicle according to claim 10, wherein the correction processing includes calculating the natural vibration frequency and the attenuation coefficient by performing frequency analysis processing on a difference between the acquired longitudinal acceleration and the model response of the longitudinal acceleration.

13. The control method for an electric vehicle according to claim 8, wherein the correction processing includes updating the dynamic characteristic when the dynamic characteristic changes more than a predetermined threshold value with respect to a previous value.

14. The control method for an electric vehicle according to claim 1, wherein the correction processing includes: calculating a natural vibration frequency and an attenuation coefficient of the longitudinal vibration component based on a longitudinal acceleration of the electric vehicle;correcting mechanical characteristics of the coupling portion or a weight of the electric vehicle or the coupled vehicle based on the natural vibration frequency and the attenuation coefficient;calculating a relative vehicle speed between the electric vehicle and the coupled vehicle based on the corrected mechanical characteristics of the coupling portion, or the corrected weight of the electric vehicle or the coupled vehicle;calculating a correction torque for the longitudinal vibration component based on the relative vehicle speed; andcorrecting the basic torque target value using the correction torque.

15. A control device for an electric vehicle that includes a motor as a drive source and a coupling portion coupled to another vehicle, and travels while towing a coupled vehicle, which is the other vehicle coupled to the coupling portion, the control device comprising: a basic torque target value calculation unit configured to calculate a basic torque target value representing a torque to be output by the motor based on a vehicle operation; anda correction processing unit configured to calculate a final torque command value, which is a final command value for the torque, by performing correction processing for reducing a longitudinal vibration component generated in the electric vehicle due to the coupled vehicle being coupled to the coupling portion on the basic torque target value, based on a dynamic characteristic of the coupling portion to which the coupled vehicle is coupled, whereinthe motor is controlled based on the final torque command value.