The present invention relates to a vehicle motion control device that controls the acceleration/deceleration of a vehicle when entering a curve and/or exiting a curve.
As a conventional vehicle motion control device that controls acceleration/deceleration during cornering (while traveling through a curve), there is known, for example, that disclosed in Patent Document 1. The object of the technique disclosed in Patent Document 1 is to provide a vehicle motion control device that clearly establishes specific principles of control timing with respect to accelerator, steering and brake manipulation, and that is able to perform motion control based thereon.
Specifically, with respect to a vehicle motion control device comprising a device that controls vehicle steering, there is provided a control means that controls the steering or longitudinal acceleration/deceleration of the vehicle using at least the vehicle's longitudinal or lateral jerk information.
In addition, in Non-Patent Document 1, principles for setting a vehicle's longitudinal acceleration/deceleration in accordance with the vehicle's lateral jerk are described.
In addition, in order to reduce the kinetic energy that the vehicle possesses, the driver may sometimes perform a deceleration operation in the zone leading up to curve entry (i.e., prior to entering a curve). As a method of automatically performing such deceleration, Patent Document 2 is known.
With respect to the vehicle motion control device disclosed in Patent Document 1, it is stated that the timing for starting or ending vehicle longitudinal acceleration control is when lateral jerk is near zero. In addition, in Non-Patent Document 1, a discussion is provided regarding basic principles for calculating vehicle longitudinal acceleration in accordance with vehicle lateral jerk that takes the acceleration/deceleration control timing of Patent Document 1 into consideration.
Considering the above from the perspective of vehicle motion, by way of example, by decelerating when lateral acceleration increases as a corner (curve) is entered, the load on the front-wheel is increased due to the fictitious force that acts on a center of gravity point having some height from the ground and the cornering stiffness of the front wheels is increased, while on the other hand the load on the rear wheels decreases, and the cornering stiffness of the rear wheels is decreased.
In addition, by accelerating when lateral acceleration decreases as a corner (curve) is exited, the load is shifted to the rear wheel, thereby stabilizing the vehicle.
In addition, in Patent Document 2, such information as the radius of a curve ahead, the distance to the curve, etc., are obtained using a navigation system, etc., a speed that results in pre-defined target lateral acceleration, in other words a target speed, is determined, and deceleration is generated in such a manner that, over the distance up to the curve, the target vehicle speed is reached from the current vehicle speed, thereby reducing the strain of driving on the driver.
Problems with known techniques regarding such deceleration methods during cornering are discussed below.
With respect to Patent Document 1 and Non-Patent Document 1, constantly imparting vehicle longitudinal acceleration in accordance with the lateral jerk generated with respect to the vehicle in such a manner may not necessarily result in control that reflects the driver's intent as s/he enters the curve.
Specifically, if deceleration is carried out in accordance with lateral jerk only (i.e., imparting deceleration because lateral jerk increases as a result of entering a curve), some drivers may experience fear due to the high speed at curve entry.
Accordingly, it is necessary to perform deceleration for the purpose of reducing kinetic energy prior to entering the curve. If, in view of the above, the methods disclosed in Patent Document 1 and Non-Patent Document 1 were to be applied to Patent Document 2, the above-mentioned deceleration amount for decelerating in accordance with lateral jerk would not be coordinated with the pre-curve entry deceleration amount, and a level difference in deceleration would occur at curve entry, thereby potentially giving the driver the impression of jerkiness in vehicle behavior.
The acceleration methods for exiting curves disclosed in Patent Document 1 and Non-Patent Document 1 are next discussed. During a steady turn, the lateral acceleration assumes a constant value, and the lateral jerk is therefore 0. When exiting a curve, since the lateral acceleration decreases in the transition zone of from a steady turn to linear travel, the lateral jerk becomes negative.
Here, although acceleration is started from when the lateral jerk is 0, the period during which acceleration is imparted only lasts while the lateral jerk is negative. When traveling the transition zone for exiting the curve, the load on the front wheels decreases due to acceleration, the load on the rear wheels increases, and the restoring yaw moment consequently increases. This, from a vehicle motion dynamics standpoint, is logical and effective in that the vehicle becomes stable.
However, from the standpoint of thereafter accelerating to a speed desired by the driver, merely accelerating only while the lateral jerk is negative may cause an unnatural feel for the driver as a result of the post-acceleration speed being too high, or conversely too low, due to there being no restrictive conditions regarding speed.
The present invention is made in order to solve such problems, and an object thereof is to provide a vehicle motion control device that enables, more safely, with less of an unnatural feel, and with an appropriate control amount, deceleration control at curve entry (deceleration control from shortly before the driver starts steering) and/or acceleration control at curve exit.
A vehicle motion control device of the present invention that solves the problems above is a vehicle motion control device that performs acceleration/deceleration control of a vehicle at curve entry and/or curve exit, the vehicle motion control device comprising: a lateral motion-coordinated acceleration/deceleration calculation means that calculates the longitudinal acceleration/deceleration of the vehicle in accordance with the lateral jerk of the vehicle; and a vehicle speed control means that calculates deceleration to be generated with respect to the vehicle before entering a curve, taking into consideration the acceleration/deceleration calculated by the lateral motion-coordinated acceleration/deceleration calculation means.
According to the present invention, since the vehicle speed control means takes the acceleration/deceleration calculated by the lateral motion-coordinated acceleration/deceleration calculation means into consideration in calculating the pre-curve entry deceleration, over-deceleration is prevented, mitigating any unnatural feel the driver may experience. Specific effects with respect to each of the claims will be described with the embodiments below. The present specification incorporates the contents of the specification and/or the drawings of JP Patent Application No. 2009-244886 from which the present application claims priority.
A configuration example of a vehicle is shown in
The revolution rates of the wheels 1a, 1b, 1c, and 1d are detected with the wheel speed sensors 2a, 2b, 2c, and 2d. Based on the revolution rates of the respective wheels detected with the wheel speed sensors 2a, 2b, 2c, and 2d of the respective wheels, the vehicle speed calculator 3 calculates vehicle speed V, which is the speed of the vehicle 0 in the travel direction. At the vehicle speed calculator 3, first, speeds Va, Vb, Vc, and Vd of the respective wheels in the travel direction are calculated by multiplying the angular speeds of the respective wheels by the turning radius of the wheels. Vehicle speed V may be the average of speeds Va, Vb, Vc, and Vd. In addition, although not shown in the drawings, a signal from a ground vehicle speed sensor using a millimeter wave radar, etc., may be taken to be vehicle speed V.
The steering angle sensor 4 detects the steering angle of the vehicle 0, and, by way of example, one that is of a generally known rotary encoder type is used. Steering angle δ detected by the steering angle sensor 4 is inputted to the vehicle motion control device 6.
The accelerator pedal opening detection sensor 5 detects the extent to which the accelerator pedal is stepped on by the driver, and it may be of a common type that turns the above-mentioned extent into an electric signal by means of, for example, a Hall element within the sensor, etc., and outputs it as a voltage.
The vehicle motion control device 6 comprises an electric circuit and a microcomputer, or just a microcomputer, and it comprises, as control elements: a lateral motion-coordinated acceleration/deceleration calculation unit 11; a vehicle speed control device 12; and an acceleration/deceleration combining unit 13. The deceleration calculated at the vehicle motion control device 6 is fed to the hydraulic brake unit 10 as a fluid pressure order value. The pre-curve entry deceleration calculation logic will be discussed later.
For the hydraulic brake unit 10, one that performs, for example, pump-up type BBW (Brake By Wire) control is used. The hydraulic brake unit 10 comprises a fluid pressure servo that feeds fluid pressure to the hydraulic brakes 8a, 8b, 8c, and 8d of the respective wheels. With respect to the brake manipulation amount by the driver, the master cylinder pressure and pedal stroke are compared with an order master cylinder pressure and order wheel cylinder pressure converted from the acceleration/deceleration order value from the vehicle motion control device 6, and the maximum value is taken to be an order value for the fluid pressure servo. The fluid pressure servo performs control in such a manner as to achieve a fluid pressure that realizes the order value, and feeds fluid pressure to the brakes 8a, 8b, 8c, and 8d of the respective wheels.
The driving force generation means 7 is a means for driving the vehicle 0, and comprises, for example, an engine (internal combustion engine), a variable gear box, and a differential. Alternatively, it may comprise a motor instead of an engine. The signal of the accelerator pedal opening detection sensor 5 is fed to an engine control unit (not shown), and the opening/closing of the throttle valve is controlled based on this information.
The stereo camera 9 comprises two monocular cameras disposed on the left and right of a frame, and realizes a function of recognizing the environment surrounding the vehicle. The number of cameras is by no means limited to two, and three or more may be provided as well. By way of example, the frame is attached near the rear-view mirror inside the vehicle cabin, and internally comprises, a CPU, RAM, ROM, etc., for processing images that have been captured.
With respect to the images captured by the left and right cameras, based on the parallax between the left and the right, distance LPC from the vehicle 0 to the entrance of a curve and minimum radius Rmin of the curve ahead are calculated from the distance to an object in the periphery of the road, and so forth. Since detailed descriptions of image processing methods can be found in known documents (e.g., JP Patent Application Publication (Kokai) No. 2008-240166, etc.), the same will be omitted herein.
A control configuration with respect to the present embodiment is shown in
The vehicle speed control device 12 calculates the deceleration that is to take place before entering the curve, and outputs it to the acceleration/deceleration combining unit 13. At the acceleration/deceleration combining unit 13, lateral motion-coordinated acceleration/deceleration Gx_dGy calculated at the lateral motion-coordinated acceleration/deceleration calculation unit 11 and deceleration Gx_PreC calculated at the vehicle speed control device 12 are combined and outputted as the final deceleration.
A control flow is described specifically using
First, in step S10, distance LPC from the vehicle 0 to curve entrance C ahead as well as curve minimum radius Rmin of the curve are calculated (curve information acquisition means). For the above, by way of example, one where the distance to the curve ahead and radius information (curve information) are transmitted to the vehicle 0 by means of a communications unit mounted on a mirror at the curve, etc., and so forth, is conceivable. However, in the present embodiment, a description is provided with respect to a method of estimating a curve based on the arrangement of lane markers and obstacles ahead using the stereo camera 9.
With respect to
The stereo camera 9 detects the distances to the left and right road edges at each point (X0, X1, X2, X3, X4 . . . ) of references points (referred to as segments) provided at regular intervals along line X extending from the vehicle's center axis parallel to the longitudinal direction of the vehicle 0.
From the above distance data, the distances between line X extending from the vehicle's center axis and the center line of the road are respectively designated y0, y1, y2, y3 . . . . The information to be calculated are distance LPC from the vehicle 0 to the transition zone, and radius Rmin of the steady turn zone.
The transition zone (zone CD) is, for ordinary roads, approximated with a clothoid curve. This may be expressed as follows, where the course of the center line of the mad is expressed with respect to a coordinate system whose origin is point C:
Assuming that radius Rmin is sufficiently greater than easement curve length LCL, the second and subsequent terms in Equations 1 and 2 may be disregarded. Accordingly, the relationship between x and y is given by the following cubic function.
[Eq. 3]
x
3=6A2y A=√{square root over (Rmin×LCL)} (3)
A in Equation 3 above is a parameter of the clothoid curve, and is expressed in terms of radius Rmin and clothoid curve length LCL. As such, this cubic function has the following relationship.
Using this relationship, distance LPC to the transition zone, and radius Rmin may be expressed by the following equations.
Here, due to the characteristics of clothoid curves, transition zone distance LCL is indeterminate. Unless the curve is actually traveled, it is impossible to detect the length of this transition zone with cameras, radars, etc., and, in practice, it would have to be estimated from the structure of the road.
By way of example, according to the Government Order on Road Design Standards, as shown in Table 1, traveled roads have respectively designated design speeds, and a transition zone distance (easement curve length) and turning radius are designated in accordance with the design speed.
While it is observed that the turning radius varies from road to mad, the variation in travel zone among roads is relatively small. As such, it is speculated that extensive use of the transition zone distances in Table 1 should pose no problems. In addition, if it is possible to obtain information on the radius of the curve ahead from map information by means of a navigation system, etc., transition zone distance LCL may be calculated, making it possible to grasp a series of characteristics of the curve with favorable precision.
In addition, in the present embodiment, the transition zone is approximated with a clothoid curve, and further approximated with a cubic function. However, there are other known methods for calculating the distance to the transition zone, the transition zone distance and the curve radius, and the method above is by no means limiting.
Thus, in the present embodiment, signals outputted by the stereo camera 9 comprise distance LPC from the vehicle 0 to curve entrance C, and minimum curve radius Rmin.
Next, in step S20, it is determined whether or not a curve lies ahead of the vehicle 0.
A schematic diagram for a curve detection determination method is shown in
The pair of dotted lines appearing above and below estimated distance Lv in
First, when distance LPC to curve entrance C falls below pre-defined threshold L_upperlim (t1), this value is taken to be an initial value, and a calculation of distance Lv to curve entrance C using the time integral of current vehicle speed V is started. Assuming that the sampling time is ts, this may be expressed by the following equation.
[Eq. 7]
L
v
=L_upperlim−V*ts (7)
Next,
[Eq. 8]
L
v
lower
<LPC<L
v
lower
→Con=Con
—
z+c (8)
[Eq. 9]
LPC<L
v
lower&LPC>Lv
where Con_z in Equations 8 and 9 above is the value of reliability Con from one unit of sampling time earlier. In other words, reliability Con is incremented by constant c as long as distance LPC lies within the range between tolerable upper limit Lv_upper and tolerable lower limit Lv_lower (t1-t2, t3-t4). However, constant c may be a fixed value, or it may be variable depending on the circumstances.
On the other hand, reliability Con retains previous value Con_z when distance LPC is outside the range between tolerable upper limit Lv_upper and tolerable lower limit Lv_lower (t2-t3, t4-t5). However, the method is not limited to retaining previous value Con_z, and may instead involve subtracting constant c.
In addition, if the state of Equation 9 mentioned above continues for a given duration (t5-t4), reliability Con is reset to 0. Thus, if reliability Con exceeds threshold Con_th, it is determined that a curve has been detected, a curve detection flag is activated as shown in
With a means that thus determines a curve, reliability Con is accumulated only while distance LPC to curve entrance C as sensed by the stereo camera 9 lies within a predetermined range relative to the distance derived through time integration of current vehicle speed V, and it is determined to be a curve when reliability Con becomes equal to or greater than a predetermined value, Con_th.
Next, a description is provided with respect to a method in step S30 of calculating distance LPC_d to curve entrance C and minimum curve radius Rmin_d which are to be ultimately outputted in the present block. It is assumed that:
[Eq. 10]
LPC
—
d=Lv Rmin
—
d=Rmin (10)
when a curve detection flag is activated (when detected) as in
Thereafter, while this curve detection flag is activated, the time constant of a first-order low-pass filter may be made greater to reduce fluctuations in minimum curve radius Rmin.
Next, in step S40, it is determined whether or not the accelerator opening detected with the accelerator pedal opening detection sensor 5 is equal to or less than a predetermined value, Apo_c. It is thus determined whether or not the driver intends to decelerate. If the accelerator opening exceeds predetermined value Apo_c, it is determined that the accelerator is stepped on and that the intention to accelerate or to maintain a certain speed is present, thereby resulting in a No. On the other hand, if it is equal to or less than predetermined value Apo_c, it is determined that the driver intends to decelerate by lifting his/her foot off the accelerator or by returning the accelerator, thereby resulting in a Yes, and the process proceeds to step S50. A determination may be made by means of a flag with predetermined value Apo_c as a threshold as in the present embodiment, or, as in
Next, in step S50, it is determined whether or not current vehicle speed V is equal to or greater than a predetermined speed, Vmin. It is assumed that no deceleration control intervention is to be performed to begin with for extremely low speeds. If current vehicle speed V is equal to or greater than predetermined speed Vmin, the result is a Yes, and the process proceeds to step S60. Again, by using a table as in the case of accelerator opening mentioned above, abrupt changes in deceleration may be suppressed.
Step S41, step S51, and step S70 will now be described. In step S41 and step S51, determinations may generally be made by methods comparable to those in step S40 and step S50.
In step S70, lateral motion-coordinated acceleration/deceleration Gx_dGy is calculated. Based on the principles of acceleration/deceleration coordinated with lateral motion disclosed in Non-Patent Document 1, lateral motion-coordinated acceleration/deceleration Gx_dGy is calculated from lateral jerk dGy and lateral acceleration Gy. An example of such a calculation method is presented below. In the present embodiment, a description will be provided with respect to a method in which lateral acceleration Gy and lateral jerk dGy are calculated from steering angle δ and vehicle speed V, and in which lateral motion-coordinated acceleration/deceleration Gx_dGy is calculated from lateral acceleration Gy and lateral jerk dGy thus calculated.
In order to calculate a lateral jerk signal from steering angle δ, it is necessary to, as with acceleration sensors, calculate lateral jerk dGy, and differentiate that signal. However, since the steering angle signal is ahead in phase relative to lateral acceleration Gy in the low frequency region, which is of significance in the context of vehicle motion, as compared to a case where lateral jerk dGy is derived by differentiating lateral acceleration Gy that occurs with respect to the vehicle 0, there is less of a response delay even when a low-pass filter of a greater time constant is applied.
A method of calculating lateral acceleration Gy and lateral jerk dGy from steering angle δ is presented. There is employed a vehicle model that outputs yaw rate r [rad/s], which is dependent on speed, with steering angle δ [deg] and vehicle speed V [m/s] as input. Yaw rate r above is expressed in terms of yaw rate gain constant Grδ(0), which does not take the second-order response delay of the vehicle 0 as given by Equation 11 below into account, and the second-order delay response with respect to steering angle δ.
With respect to Equation 11 above, the terms Tr, ζ, and con are parameters unique to the vehicle, and are values that are pre-identified empirically. Next, from gain constant Grδ(0), lateral acceleration Gy is given by Equation 12 below.
[Eq. 12]
G
y
=V(dβ+Gδr×δ) (12)
The term dβ in Equation 12 above is the rate of change in side slip angle. However, for motion that is within the linear region of tire force, dβ may be substantially disregarded as being negligible.
Next, lateral acceleration Gy that has been calculated undergoes discrete differentiation and is passed through a low-pass characteristics filter to obtain lateral jerk dGy. Time constant Tlpf of the low-pass characteristics filter in this case takes the second-order response delay mentioned earlier into account. In addition, in order to coordinate the phases, lateral acceleration Gy that has been passed through the same low-pass characteristics filter of time constant Tlpf is used.
Using lateral acceleration Gy and lateral jerk dGy thus calculated, lateral motion-coordinated acceleration/deceleration Gx_dGy of the vehicle 0 is calculated in accordance with Equation 13 below.
Equation 13 above basically multiplies lateral jerk dGy by gain Cxy to obtain a value to which a first-order delay is imparted. To further generalize the above, acceleration/deceleration that is coordinated with lateral motion and produces less of an unnatural feel may also be achieved by an embodiment in which lateral jerk dGy is multiplied by proportionality coefficient Cxy, as represented by Equation (14) below.
[Eq. 14]
G
x
dGy
=−C
xy
·|dG
y| (14)
Proportionality coefficient Cxy in Equation 14 above may be varied based on speed V, the range of lateral acceleration Gy, side slip condition, etc.
In the present embodiment, a description has been provided taking as an example a case where calculations are performed using detection values of the various sensors shown in
In addition, lateral jerk dGy calculated from steering angle δ may be construed as lateral jerk dGy intended by the driver, and there is a discrepancy between actual lateral jerk dGy and that calculated from steering angle δ. As such, two values of lateral motion-coordinated acceleration/deceleration Gx_dGy may be calculated respectively using both lateral jerk dGy calculated from steering angle δ (feed forward) and actual lateral jerk dGy (feedback), and the two may be combined. Thus, in step S70, lateral motion-coordinated acceleration/deceleration Gx_dGy corresponding to lateral jerk dGy is calculated.
Next, with respect to step S60, a method of calculating estimated lateral jerk Gx_dGypre in transition zone CD of the curve is described. As in
When lateral acceleration Gy that occurs with the transition to curve zone CDE is plotted on a graph, the result would resemble
Thus, as indicated in Equation 14, Gx_dGypre, the deceleration estimated to occur in this zone (estimated lateral motion-coordinated acceleration/deceleration), may be expressed by Equation 16 below.
In ideal transition zone CD formed of a clothoid curve, if the driver increases steering angle δ at a constant steering speed, lateral acceleration Gy would increase at a constant rate. Accordingly, it is speculated that no significant sense of unnaturalness would be experienced even if lateral jerk dGy, which is the rate at which lateral acceleration Gy increases, were approximated with a linear function as in
Next, with respect to step S80, a description is provided regarding a method of calculating, before entering a curve, pre-curve entry deceleration Gx_preC, which takes estimated lateral motion-coordinated acceleration/deceleration Gx_dGypre into account. A state where the driver lifts his/her foot off the accelerator at point B shown in
Thereafter, in transition zone CD, deceleration occurs based on estimated lateral motion-coordinated acceleration/deceleration Gx_dGypre, as a result of which the vehicle speed at point D becomes Vmin. This may be expressed in an equation as follows.
[Eq. 18]
V
ent
=V
min
+C
xy
G
y
max (18)
In addition, vehicle speed Vmin may be expressed as follows using minimum curve radius Rmin and maximum lateral acceleration (target lateral acceleration) Gy_max.
The following constraint is imposed on pre-curve entry deceleration Gx_preC that occurs here and maximum lateral acceleration Gx_max at point D.
[Eq. 20]
G
y
max
=αG
x
preC (20)
Thus, it is possible to define the magnitudes of the deceleration to be generated before entering the curve (pre-curve entry deceleration Gx_preC) and lateral acceleration Gy during steady turning. By way of example, when α is 1, pre-curve entry deceleration Gx_preC and maximum lateral acceleration Gx_max become equal. When it is 10, pre-curve entry deceleration Gx_preC, which is the deceleration that occurs before entering the curve, becomes smaller, and because it becomes smaller, speed Vent upon reaching point C becomes greater and the lateral jerk that consequently occurs also increases. Thus, lateral motion-coordinated acceleration/deceleration Gx_dGy also becomes greater as per Equation 14.
With respect to the above, by altering a by manipulating a switch inside the vehicle, for example, the driver would be able to vary the magnitudes of the pre-curve entry deceleration and the deceleration in the transition zone. Since preferences regarding the magnitude of pre-curve entry deceleration vary from driver to driver and cannot be determined uniquely, the present method is believed to be effective.
Further, from another perspective, for example, by obtaining external environment information, such as lane width, narrowness of view, etc., in addition to distance LPC to curve entrance C and curve minimum radius Rmin, as numerical information by means of the stereo camera 9, the amount of deceleration to occur beforehand may be increased, that is, a may be reduced, if the road is narrow, thereby preventing traveling a curve at too high a speed and reducing the driver's anxiety. Further, using navigation information, if, based on past travel data, the road currently traveled is a road that has been traveled in the past, its level of familiarity may be quantified, and if it is determined that the driver is familiar with it, the pre-curve entry deceleration amount may be reduced, that is, α may be increased, thereby presumably making it less likely that the driver would find it sluggish.
Through Equations 18-20, pre-curve entry deceleration Gx_preC that is to occur shortly before entering the curve may be expressed as
[Eq. 21]
α2Cxy2A4+2αCxy√{square root over (αR)}A3+(αR−2LPC)A2−V02=0 A=√{square root over (Gx
Of the solutions to the biquadratic equation above, the positive solution is used as pre-curve entry deceleration Gx_preC. In addition, this does not have to be analytically solved, and a solution may instead be derived from an equation that has been simplified by approximation.
Next, with respect to step S90, a description is provided regarding a method of combining decelerations. In step S90, acceleration/deceleration order value Gx_order to be ultimately outputted is calculated based on pre-curve entry deceleration Gx_preC and lateral motion-coordinated acceleration/deceleration Gx_dGy.
Changes in deceleration across points A through D are shown in
Pre-curve entry deceleration Gx_preC given by Equation 21 above, and lateral motion-coordinated acceleration/deceleration Gx_dGy given by Equation 14 above vary as shown in
In this case, it is assumed that acceleration/deceleration order value Gx_order is passed through the first-order low-pass filter, etc., of pre-curve entry deceleration Gx_preC. Instead of just a low-pass filter, the driver's accelerator manipulation (accelerator opening speed) from point A to point B may also be taken into consideration where, if the accelerator pedal is released relatively quickly, deceleration may be made to rise slightly faster, whereas if the accelerator is returned slowly, deceleration may be made to rise slightly more slowly.
In addition, if variations were to occur in pre-curve entry deceleration Gx_preC due to detection variations by the stereo camera 9, the driver would experience an unnatural feel due to fluctuations in deceleration. As such, as shown in
Next, from point C, lateral motion-coordinated acceleration/deceleration Gx_dGy on the deceleration side begins to rise. It would ideally rise instantaneously to a value equal to pre-curve entry deceleration Gx_preC, but since there exists, at CC′ during which the driver's steering speed becomes constant, a zone in which lateral jerk dGy increases, it changes as shown in
Specifically, the fluctuation in deceleration that occurs when lateral motion-coordinated acceleration/deceleration Gx_dGy rises again after pre-curve entry deceleration Gx_preC has become 0 translates into a sense of unnaturalness for the driver. In addition, a change in the longitudinal acceleration of the vehicle 0 in zone CD where lateral acceleration Gy increases also translates into a sense of unnaturalness for the driver. Accordingly, in zone CC′, deceleration (acceleration/deceleration order value Gx_order) is so controlled as to maintain a constant value.
However, thus maintaining deceleration at a constant value affects lateral motion-coordinated acceleration/deceleration Gx_dGy in such a manner that it becomes lower as compared to a case where deceleration is not maintained at a constant value. However, since zone CC′ is short in duration, it is negligible for practical purposes.
Since lateral jerk dGy thereafter decreases as point D is approached, lateral motion-coordinated acceleration/deceleration Gx_dGy decreases. By way of example, as shown in
In addition, if, for example, lateral motion-coordinated acceleration/deceleration Gx_dGy exceeds acceleration/deceleration order value Gx_order, acceleration/deceleration order value Gx_order may be made equal to lateral motion-coordinated acceleration/deceleration Gx_dGy as shown in
Through such a sequence of control, acceleration/deceleration order value Gx_order may be varied smoothly from AC before entering the curve across transition zone CD, albeit with some fluctuation, and the sense of unnaturalness caused by fluctuations in deceleration may be mitigated.
In addition, although, in the present embodiment, it is assumed that output is made to a brake actuator as an actuator (an acceleration/deceleration means) that realizes the deceleration of acceleration/deceleration order value Gx_order, this is by no means limiting, and even with respect to common hybrid vehicles comprising, as vehicle components, a motor and a brake actuator, it may be realized by distributing deceleration between the motor and the brake actuator.
In addition, by further using engine braking and an automatic transmission (AT) or a continuously variable gear ratio transmission (CVT), subtracting the deceleration attainable through engine braking from acceleration/deceleration order value Gx_order, and allotting the remaining deceleration to the brake actuator, brake pad wear may be reduced. In this case, although the command value for lateral motion-coordinated acceleration/deceleration Gx_dGy may sometimes be faster than the response speed of engine braking, delicate deceleration may be achieved by a brake actuator that is faster than the response of engine braking.
Embodiment 2 is next described. In Embodiment 2, descriptions are provided with respect to the coexistence of driver operation and lateral motion-coordinated acceleration/deceleration Gx_dGy at the time of curve entry, as well as to acceleration control at the time of curve exit.
A configuration example of a vehicle is shown in
The driving force generation means 7 is an internal combustion engine in the present embodiment. The two front wheels, 1a and 1b, are driven via a transmission and a differential. The generator 14 directly connected to the front axle is driven to rotate by means of the power obtained from the engine 7. The electric power generated at this point becomes electric power for driving the battery 15, and is fed to the rear-wheel motor 16 via the differential. With respect to the above, orders are fed to the various elements, e.g., the engine, the generator, the motor, the battery, etc., by means of a hybrid controller (not shown) to perform the desired operation.
A brake pedal 17 quantifies the driver's brake manipulation amount by means of a stroke sensor, etc., and feeds it to the vehicle motion control device 6. In the present embodiment, the vehicle motion control device 6 outputs to the hybrid controller the driving forces and brake order values for the respective wheels, thereby enabling driving by the motor of the front wheels 1a and 1b and the engine, regeneration by the front-wheel motor only, driving and regeneration by the rear-wheel motor 16, and braking by the hydraulic brake actuators 8a through 8d. With such a configuration, it is possible to quantify the driver's brake pedal manipulation and to distribute the braking force among the motor and the hydraulic brakes.
A control configuration with respect to the present embodiment is shown in
In the present embodiment, a scene where the driver enters a curve with his/her own brake manipulation is assumed. With the technique disclosed in Non-Patent Document 1, if the driver steps on the brake pedal from shortly before entering the curve (zone AC) and terminates his/her brake manipulation at transition zone CD, there occurs a level difference between the deceleration based on the driver's brake order value and lateral motion-coordinated acceleration/deceleration Gx_dGy as shown in
As such, if, as in
In addition, if, as shown in
When controlled thus, the driver's brake order value and the acceleration/deceleration order value by the vehicle motion control device 6 become continuous, thereby mitigating the unnatural feel experienced by the driver.
A control configuration with respect to the present embodiment is shown in
In the present embodiment, as shown in
Patent Document 1 and Non-Patent Document 1 disclose that it is empirically known that gain Cxy in Equation 14 assumes, as a fixed value, a value of 0.3 to 0.5. However, this is restricted to cases where lateral motion-coordinated acceleration/deceleration Gx_dGy is calculated as a negative value, that is, as a deceleration order value, and it cannot be ascertained when accelerating.
Accordingly, since the constraint of reaching a predetermined speed is absent when the transition zone, as a road shape, is short, there were cases where the acceleration control terminated without resulting in sufficient acceleration, or cases where the speed was conversely too high due to excessive acceleration.
As such, in the present embodiment, a description is provided regarding a method in which lateral jerk dGy that occurs in transition zone GH ahead is estimated from a state where steady turn zone FG is being traveled, speed Vout that is ultimately to be reached at point H, which is the curve exit, is defined, and gain Cxy_accel at that point in time is determined.
The change in lateral acceleration over zone FI is shown in
Easement curve zone GH is a transition zone between the curve's steady turn zone FG and straight zone HI, and as the vehicle 0 travels towards the curve exit, lateral acceleration Gy acting thereon decreases. The rate by which lateral acceleration Gy thus decreases represents lateral jerk dGy and may be estimated as follows using a linear function.
Although Equation 22 differs from Equation 15 in sign, it may be expressed with the same equation. As stated in Embodiment 1, in an ideal transition zone formed of a clothoid curve, if the driver releases steering at a constant steering speed, lateral acceleration Gy would decrease at a constant rate. Accordingly, it is speculated that a method of approximation that uses a linear function as in
Based on Equation 22, estimated lateral motion-coordinated acceleration/deceleration Gx_dGypre estimated to occur in transition zone GH is expressed by the following equation.
This estimated lateral motion-coordinated acceleration/deceleration Gx_dGypre is outputted to the vehicle speed control device 12 as the output of the lateral motion-coordinated acceleration/deceleration calculation unit 11. Next, the vehicle 0 is accelerated from vehicle speed Vmin to vehicle speed Vout by estimated lateral motion-coordinated acceleration/deceleration Gx_dGypre. This may be expressed in an equation as follows.
[Eq. 24]
V
out
=V
min
+C
xy
accel
G
y
max (24)
Here, if vehicle speed Vent at point C at which deceleration based on lateral motion-coordinated acceleration/deceleration Gx_dGy is started in
In other words, by using lateral acceleration Gx_max, which has reached a maximum value while turning, vehicle speed Vmin at that point in time, and vehicle speed Vent at point C, it is possible to determine gain Cxy_accel for reaching vehicle speed Vout (Vent) at point H, which is the curve exit. Thus, the acceleration outputted by the lateral motion-coordinated acceleration/deceleration calculation unit 11 is given by
[Eq. 26]
G
x
dGy
=C
x
accel
·|dG
y (26)
and is outputted to the engine control unit.
Through such control, by simply traveling through transition zone GH by stepping on the accelerator as it had been to maintain vehicle speed Vmin, it is possible to restore the vehicle speed from vehicle speed Vout to vehicle speed Vent by the time point H (curve end point) is reached. Thus, the driver operation of having to change the speed in accordance with the travel situation may be mitigated.
As such, as in acceleration/deceleration order value Gx_order in
An interface of the vehicle motion control device 6 is presented in
Next, by turning the dial, the driver is able to select, as desired, one of safety mode (Sd), normal mode (No), and sport mode (Sp). These modes respectively vary a in Equation 20. By way of example, it may be assumed that α=1 in safety mode, that α=2 in normal mode, and that α=3 in sport mode.
It is thus possible to adjust the deceleration to be generated shortly before entering a curve. In so doing, as in
In so doing, if acceleration/deceleration order value Gx_order is negative, an indication of orange and deceleration control in effect is made, and the system On indication is returned to once acceleration/deceleration order value Gx_order ceases to be negative. In addition, if, in a system On indication state, acceleration/deceleration order value Gx_order becomes positive, an indication of light blue and acceleration control in effect is made. Once acceleration/deceleration order value Gx_order ceases to be positive, the system On indication is returned to.
The interface described above is merely an example. By way of example, instead of a push-button type dial, mode switching made be performed via voice recognition, or various switches may be aggregated on the steering wheel.
Examples of yaw moments that are generated with respect to the vehicle 0 when zone GH is traveled by accelerating in accordance with Equation 26 are shown in
The acceleration order value given by Equation 26 increases the negative yaw moment, commonly referred to as the restoring yaw moment, in order to bring the yaw rate that occurs while turning to 0 (a straight travel state) over the course of transition zone GH. If no acceleration takes place, this is caused based on the lateral force difference between the front wheels and the rear wheels and on the center of gravity position. However, if acceleration does take place in this zone, the load shifts from the front wheels to the rear wheels, which causes the lateral force difference to become even greater, and the restoring yaw moment that occurs at this point becomes even greater. Accordingly, it is possible to return to a straight travel state more quickly.
However, as shown in
Next, Embodiment 3 is described. In Embodiment 3, descriptions are provided regarding an embodiment where, with respect to the vehicle speed control device 12, control that holds the host vehicle speed constant (hereinafter referred to as cruise control) is combined.
The vehicle motion control device 6 of the present embodiment comprises the lateral motion-coordinated acceleration/deceleration calculation unit 11, the vehicle speed control device 12, and the acceleration/deceleration combining unit 13. Based on the vehicle speed calculated by the vehicle speed calculator 3, the vehicle speed control device 12 performs a torque order with respect to the engine control unit (not shown) to maintain that vehicle speed. The engine control unit calculates the throttle opening that would attain the ordered torque with the current engine revolution rate, and controls the throttle valve.
The cruise control of the present embodiment is described using the graphs shown in
c) shows accelerator pedal opening.
On the condition that the driver is manipulating neither the accelerator pedal nor the brake pedal at this point, the vehicle speed control device 12 takes the vehicle speed at the time (ta) at which flag f_CC_On changed from 0 to 1 to be a target vehicle speed, performs feedback control of the current vehicle speed and the target vehicle speed to maintain this, calculates a cruise control order torque, and feeds it to the engine control unit.
The vehicle speed control device 12 constantly compares the driving force requested by the driver and the cruise control order torque, and outputs the greater of the two. Accordingly, if the driver next steps on the accelerator from this state (tb), and the vehicle speed control device reads the accelerator opening and converts it into the driver's requested torque to produce a result that is greater than the cruise control order torque for maintaining the vehicle speed, that driver's requested torque is outputted to the engine control unit. Thus, the vehicle speed increases.
Next, if the driver eases the accelerator pedal (tc), and the driver's requested torque falls below the cruise control order torque for maintaining the vehicle speed, the vehicle speed at that point is memorized, and feedback control for the current host vehicle speed is performed with that speed as the target vehicle speed.
Next, it is assumed that the driver steps on the brake (td). Since this causes the driver's requested torque to become negative, the driver's requested torque is achieved through engine braking, hydraulic brakes, the motor's regenerative torque, etc. Meanwhile, the vehicle speed decreases.
Next, it is assumed that the driver releases the brake (te). As a result, the vehicle speed at the time at which the brake was released is taken to be the target vehicle speed, feedback control is performed in such a manner as to maintain that vehicle speed, and an output to the engine control unit is made as a cruise control order torque.
If the driver then turns the cruise control switch Off, control for maintaining the vehicle speed ceases to be performed, and the vehicle speed decreases in accordance with engine braking. Accordingly, when the cruise control switch is On, the torque for keeping the current vehicle speed constant and the driver's requested torque, which is calculated based on at least one of the driver's accelerator manipulation, brake manipulation, Equation 14, and Equation 26, are compared, and the greater of the two is outputted. In addition, if the driver's requested torque is negative (e.g., when braking, etc.), the driver's requested torque holds priority. With such an operation, it becomes possible to mitigate the operational load on the driver.
By way of example, a scene is assumed where a curve is entered with the cruise control switch turned On. The driver performs no accelerator or brake manipulations, and enters the curve with the speed held constant by cruise control. As shown in
In the present embodiment, deceleration intervenes when this flag is activated, and deceleration takes place in accordance with the decelerations of Equation 21 and Equation 14. Thereafter, a steady turn is performed while maintaining the speed from when point D in
Next, when the curve is exited and a transition is made to a straight line zone, to provide a description using
It is noted that the present invention is by no means limited to the various embodiments discussed above, and that various modifications are possible within a scope that does not depart from the spirit of the present invention.
With respect to a vehicle motion control device of the present invention, a vehicle speed control device comprises a curve detection means that detects a curve ahead of the vehicle, and if the distance to the curve entrance detected by the curve detection means is within a predetermined range relative to the estimated distance to the curve calculated through time integration of the host vehicle speed with the distance to the curve entrance at a given time as an initial value, the reliability of curve detection is made to be greater as compared to when it falls outside of the predetermined range.
In addition, with respect to a vehicle motion control device of the present invention, the vehicle speed control device maintains the reliability of curve detection when the distance to the curve entrance detected by the curve detection means falls outside of the predetermined range relative to the estimated distance to the curve calculated through time integration of the host vehicle speed with the distance to the curve entrance at a given time as an initial value.
Furthermore, with respect to a vehicle motion control device of the present invention, the vehicle speed control device accumulates the reliability of curve detection from when the distance to the curve entrance detected by the curve detection means lies within the predetermined range relative to the estimated distance to the curve calculated through time integration of the host vehicle speed with the distance to the curve entrance at a given time as an initial value, and determines that a curve lies ahead if the reliability reaches or exceeds a pre-defined value.
Furthermore, with respect to a vehicle motion control device of the present invention, the vehicle speed control device calculates a driver's requested braking/driving torque, which is converted from at least one of the vehicle's accelerator opening, brake manipulation amount, and lateral motion-coordinated acceleration/deceleration, and a given vehicle speed torque that is required to keep the current vehicle speed constant, and outputs the greater of the absolute values of the two if both torques are of the same sign, or the driver's requested braking/driving force if they are of different signs.
Furthermore, while it is determined that there lies a curve, a vehicle motion control device of the present invention makes a correction in accordance with at least one of maintaining the curve radius detected by the curve detection means, increasing the time constant for when a first-order low-pass filter is passed, and decreasing the tolerable increase/decrease range with respect to time.
With respect to a vehicle motion control device of the present invention, based on curve information and vehicle speed, a lateral motion-coordinated acceleration/deceleration calculation means calculates the maximum lateral acceleration that acts on the vehicle while traveling through a curve, and calculates the estimated lateral jerk based on that maximum lateral acceleration.
This vehicle speed may be defined as the vehicle speed when the amount by which the accelerator pedal is stepped on prior to a curve becomes equal to or less than a pre-defined threshold. In addition, the vehicle speed may be defined as the vehicle speed at the moment when it is determined by the curve detection means that a curve lies ahead of the vehicle. In addition, the vehicle speed may be defined as the speed when the calculation of the lateral motion-coordinated acceleration/deceleration is started by the lateral motion-coordinated acceleration/deceleration calculation device.
Number | Date | Country | Kind |
---|---|---|---|
2009-244886 | Oct 2009 | JP | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/JP2010/063515 | 8/10/2010 | WO | 00 | 4/20/2012 |