Machine position control device

Abstract

An accurate load position control is made possible even when rigidity of a load drive system using a motor is relatively low. Load position signal xl is a present-position measurement value of a load, and after compensation in response to a phase delay thereof has been performed by a stability compensation circuit, the high-frequency portion thereof is taken as a control-target position signal xfb by replacing, in a position-signal combination circuit, a motor position signal xm as a present-position measurement value of a motor, and then the control-target position signal xfb is fed back to a position control circuit. Thereby, a torque command signal indicating a torque target value for driving the load is outputted.

Description

TECHNICAL FIELD

The present invention relates to machine position control devices such as machine tools and component mounting machines, for controlling positions of machine systems, which are driven using an actuator such as a motor.

BACKGROUND ART

A conventional machine position control device has been configured to enhance stability of its control system using, as a feedback signal for the position controller, a signal obtained by summing up signals obtained by low-pass filtering a load-position signal as a load-position detection value, and by high-pass filtering a motor-position signal as a motor-position detection value: a motor-position signal with no phase delay in a frequency band over its resonance frequency where the phase delay of the load-position signal becomes significant, because a driving system using a motor has a limited rigidity (for example, refer to Patent Document 1).

Moreover, in another conventional machine position control device, a pre-compensator is provided therein which configuration enables high-accurate control with a deformation error along moving directions of a machine system being compensated, by adding to an input parameter, a signal, as a feed-forward compensation value, obtained by multiplying by a gain the second order differential value of the input parameter (for example, refer to Patent Document 2).

[Patent Document 1] Japanese Laid-Open Patent Publication No. 2004-334772 (FIG. 1)

[Patent Document 2] Japanese Laid-Open Patent Publication No. 1999-184529 (FIG. 3 and FIG. 4)

DISCLOSURE OF THE INVENTION

Problems to be Solved by the Invention

In the configuration in which the signal obtained by summing the signal having been obtained by low-pass filtering the load position signal and the signal having been obtained by high-pass filtering the motor position signal is used as the feedback signal into the position controller, in order to improve the control accuracy of the load position, the filter frequencies of the high-pass filter and the low-pass filter are necessary to be substantially increased. However, in a case in which the rigidity of the load driving-system with the motor is relatively low, if the filter frequency is increased, the control system becomes unstable; therefore, because the filter frequency could not be sufficiently increased, a problem has occurred that the load position is difficult to be accurately controlled.

Moreover, in the configuration in which the pre-compensator is provided, because the feed-forward compensation value is calculated based on the second order differential value of the input parameter based on the position command signal, and is summed, varying of a torque command signal corresponding to that of the position command signal becomes steep, and impact applied to a control target increases; therefore, the gain of the position controller cannot be sufficiently increased. As a result, accurate position control of the load is difficult to be realized; moreover, a problem has occurred that vibration generated when an external disturbance is exerted to the control target cannot be suppressed.

Means for Solving the Problem

A machine position control device according to the present invention includes a position-speed control circuit in which a torque command signal representing a torque target value at which a motor drives a load is calculated by summing a position command signal representing a load-position target value, and feedback signals of a motor position signal representing a present position of the motor, a motor speed signal representing a present speed of the motor, and a control-target position signal as reference information related to the present positions of the motor and the load. The control-target position signal is configured in such a way that a signal including a low-frequency component of a compensated load-position signal obtained by compensating the phase delay thereof in a stability compensation circuit, and a signal including a high-frequency component of a motor position signal are combined in a position-signal combination circuit, based on a transfer function for advancing the phase of a load position signal that is the measurement value of the load present-position.

Advantageous Effect of the Invention

According to the present invention, even when the rigidity of the load drive system using the motor is relatively low, by performing the load position control in which the load position signal is fed back, as stable as the semi-closed control in which only the motor position signal is fed back, an accurate load position control can be realized.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 is a block diagram illustrating a machine position control device according to Embodiment 1 of the present invention;

FIG. 2 is views representing a frequency response of a control target;

FIG. 3 is a block diagram illustrating a machine position control device according to a semi-closed control system;

FIG. 4 is a block diagram illustrating an internal configuration of a position-signal combination circuit;

FIG. 5 is a block diagram illustrating an internal configuration of a stability compensation circuit;

FIG. 6 is a block diagram illustrating an internal configuration of a damping compensation circuit;

FIG. 7 is a block diagram illustrating a machine position control device according to Embodiment 3 of the present invention; and

FIG. 8 is a block diagram illustrating a machine position control device according to Embodiment 4 of the present invention.

EXPLANATION OF SYMBOLS

• 10 Control target
• 20 Load
• 30 Motor
• 80, 80a Stability compensation circuit as stability compensation means
• 110 Position-speed control circuit as position-speed control means
• 140 Damping compensation circuit as damping compensation means

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1

FIG. 1 is a block diagram illustrating a machine position control device according to Embodiment 1 of the present invention.

A control target 10 includes a motor 30 for driving a load 20, in which the motor 30 drives the load 20 through a torque transmission mechanism 40 such as a timing belt and a ball screw, and the torque τ_m of the motor 30 is controlled so as to be agree with a torque command signal τ_r by a torque control circuit 50.

Moreover, the present position of the motor 30 is detected by a motor position detector 60 such as an encoder attached to the motor 30, and is outputted as a motor position signal x_m, while the present position of the load 20 is detected by a load position detector 70 such as a linear scale attached to the load 20, and is outputted as a load position signal x_l.

A stability compensation circuit 80 receives the load position signal x_l as input, and outputs a compensated load-position signal x_lc in which the phase delay of the load position signal x_l is compensated; and a position-signal combination circuit 90 receives as input the compensated load-position signal x_lc and the motor position signal x_m, and outputs a control-target position signal x_fb that is a feedback signal related to the position of the motor and the load.

A speed calculation circuit 100 receives the motor position signal x_m as input, and outputs a motor speed signal v_m that represents the present value of the motor speed.

A position-speed control circuit 110 is configured of a position gain circuit 120 that receives as input a position command signal x_r and the control-target position signal x_fb, and outputs a speed command v_r as a speed target value, and of a speed PI-control circuit 130 that receives as input the speed command v_r and the motor speed signal v_m, and outputs a basic control-torque signal τ_b to be a basis for calculating the torque command signal τ_r.

A damping compensation circuit 140 receives as input the position command signal x_r, the motor position signal x_m, and the load position signal x_l, and outputs, based on a damping adjustment parameter α set from the outside, a damping-compensation torque signal τ_c that compensates the basic control-torque signal τ_b. A signal obtained by adding the damping-compensation torque signal τ_c to the basic control-torque signal τ_b agrees with the torque command signal τ_r.

Next, an operation is explained.

The position gain circuit 120 outputs as the speed command v_r a signal obtained by multiplying by a position gain k_p a deviation between the position command signal x_r and the control-target position signal x_fb. That is, calculation of the following equation is performed.[Equation 1]v_r=k_p(x_r−x_fb)  (1)

Next, as expressed by the following equation, the speed calculation circuit 100 outputs the motor speed signal v_m by differentiating the motor position signal x_m.[Equation 2]v_m=s·x_m  (2)

Next, the speed PI-control circuit 130 receives as input the speed command v_r and the motor speed signal v_m, and outputs, using a speed gain k_v and a speed-integration gain ω_vi, the basic control-torque signal τ_b according to PI (proportional integral) calculation expressed by the following equation.

$\begin{array}{cc}\left[\mathrm{Equation}3\right]& \\ {\tau }_b-\frac{k_v\left(s+{\omega }_{\mathrm{vi}}\right)}{s}\left(v_r-v_m\right)& \left(3\right)\end{array}$

Next, characteristics of the control target 10 are explained.

When mechanical rigidity of the control target 10 is relatively low, the control target 10 has the characteristics of a mechanical resonance whose frequency is relatively low (from several Hz to several-ten Hz). Focusing on the lowest-frequency mechanical-resonance characteristic, the characteristic of the control target 10 can be approximated by a two-inertia system in which the motor 30 and the load 20 are connected with the torque transmission mechanism 40 as a spring. Therefore, if the responsiveness of the torque control circuit 50 is sufficiently fast, a transfer function G_p(s) from the torque command signal τ_r to the motor position signal x_m, and a transfer function G_l(s) from that to the load position signal x_l are respectively expressed as follows:

$\begin{array}{cc}\left[\mathrm{Equation}4\right]& \\ \frac{x_m}{{\tau }_r}=G_p\left(s\right)=\frac{{\omega }_z^{-2}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(4\right)\\ \left[\mathrm{Equation}5\right]& \\ \frac{x_1}{{\tau }_r}=G_1\left(s\right)=\frac{1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(5\right)\end{array}$ where J denotes all inertia effects of the control target 10, ω_z denotes the anti-resonance frequency, and ω_p denotes the resonance frequency.

Here, the transfer function G_p(s) from the torque command signal τ_r to the motor position signal x_m has complex zeroes (anti-resonance points) z₀ corresponding to the anti-resonance frequency ω_z.[Equation 6]z₀=±j·ω_z  (6)

Frequency response of the transfer functions of the control target 10, expressed by Equation 4 and Equation 5, is represented in FIG. 2. FIG. 2 represents that, regarding the transfer function G_p(s) from the torque command signal τ_r to the motor position signal x_m, the phase is never delayed exceeding −180 degrees, but, regarding the transfer function G_l(s) to the load position signal x_l, the phase is significantly delayed at the resonance frequency ω_p.

Next, in order to explain operations of the stability compensation circuit 80 and the position-signal combination circuit 90, a semi-closed control system in which the feedback with respect to the position of the load 20 is not utilized is explained, which is a control system most widely used in a case in which a machine system is driven using a motor.

FIG. 3 is a block diagram illustrating a configuration of a semi-closed control system, in which neither the load position detector 70 nor the load position signal x_l is provided compared to the configuration in FIG. 1. Moreover, the position-signal combination circuit 90, the stability compensation circuit 80, and the damping compensation circuit 140 are not provided; thus, the motor position signal x_m is directly inputted as the control-target position signal x_fb into the position gain circuit 120.

Because in the semi-closed control system represented in FIG. 3, the load position signal x_l is not fed back, if deformation of the torque transmission mechanism 40 is included therein, the position of the load 20 cannot be exactly controlled. However, the position gain k_p of the position gain circuit can be set relatively high with the stability of the control system being maintained; therefore, the system has a characteristic that the response for controlling the motor position signal x_m can be increased.

Moreover, in the semi-closed control system, an open-loop transfer function L(s) obtained by opening the entire control loop at the position of the torque command signal τ_r (also called as a single loop transfer function; hereinafter, simply called as an open-loop transfer function) can be expressed by the following equation using the speed gain k_v and the speed-integration gain ω_vi.

$\begin{array}{cc}\left[\mathrm{Equation}7\right]& \\ \begin{array}{c}L\left(s\right)=\frac{k_v\left\{s^2+\left(k_p+{\omega }_{\mathrm{vi}}\right)s+k_p{\omega }_{\mathrm{vi}}\right\}}{s}{G}_p\left(s\right)\\ =-\frac{k_v\left\{s^2+\left(k_p+{\omega }_{\mathrm{vi}}\right)s+k_p{\omega }_{\mathrm{vi}}\right\}}{s}\frac{{\omega }_z^{-2}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\\ =-\frac{k_v\left(s+k_p\right)\left(s+{\omega }_{\mathrm{vi}}\right)}{s}\frac{{\omega }_z^{-2}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\end{array}& \left(7\right)\end{array}$

In the open-loop transfer function L(s) of the semi-closed control system, the transfer function G_p(s), of the control target 10, from the torque command signal τ_r to the motor position signal x_m is included as an element. Therefore, the anti-resonance points z₀ included in G_p(s) are intactly included as zeroes of the open-loop transfer function. While, zeroes of the open-loop transfer function, except for the anti-resonance points, are real number zeroes of −ω_vi and −k_p set in the speed PI-control circuit 130 and the position gain circuit 120.

On the other hand, regarding the semi-closed control system, by performing the feedback of the load position signal x_l, the position of the load 20 can also be considered to be exactly controlled despite deformation of the torque transmission mechanism 40. However, if the load position signal x_l is used without modification, the control in a range higher than a predetermined frequency becomes unstable due to the effect of phase delay caused by low mechanical rigidity of the control target 10; therefore, a configuration has been considered as disclosed in Patent Document 1, in which feedback of the motor position signal x_m is performed in a frequency range higher than a predetermined one.

The configuration is equivalent to that in FIG. 1 in which the damping-compensation signal τ_c is set to nil by removing the damping compensation circuit 140, and the load position signal x_l, instead of the compensated load-position signal x_lc, is directly inputted into the position-signal combination circuit 90 by removing the stability compensation circuit 80. That is, the feedback of a signal, obtained by combining the load position signal xl and the motor position signal xm in the position-signal combination circuit 90, into the position gain circuit 120 is performed.

FIG. 4 is a block diagram illustrating an internal configuration of the position-signal combination circuit 90.

A motor-position filter 91 receives the motor position signal x_m as input, and outputs a filtered signal obtained by using a high-pass filter F_m(s) whose filter frequency is ω_f. While, a load-position filter 92 receives the compensated load-position signal x_lc as input, and outputs a filtered signal obtained by using a low-pass filter F_l(s) whose filter frequency is the same as ω_f the motor-position filter 91.

Thereby, a summed signal obtained by summing the output from the motor-position filter 91 and that from the load-position filter 92 is outputted from the position-signal combination circuit 90 as the control-target position signal x_fb.

That is, the position-signal combination circuit 90 performs calculation expressed by the following equation.

$\begin{array}{cc}\left[\mathrm{Equation}8\right]& \\ \begin{array}{c}{x}_{\mathrm{fb}}=F_m\left(s\right)x_m+F_1\left(s\right)x_{1c}\\ =\frac{s}{s+{\omega }_f}{x}_m+\frac{{\omega }_f}{s+{\omega }_f}{x}_{1c}\end{array}& \left(8\right)\end{array}$

That is, the position-signal combination circuit 90 synthesizes the control-target position signal x_fb from the low-frequency component of the compensated load-position signal x_lc and the high-frequency component of the motor position signal x_m. Thus, the circuit is configured in such a way that the higher the filter frequency ω_f, the more the ratio of utilizing the compensated load-position signal x_lc than the motor position signal x_m is increased.

However, the control-target position signal x_fb is generated by synthesizing the frequency components of the motor position signal x_m and the load position signal x_l, and the load position signal x_l is excessively included in the control-target position signal x_fb at a frequency lower than the filter frequency ω_f. Therefore, in order to improve the control accuracy of the load position signal x_l, not only by sufficiently increasing the position gain k_p of the position gain circuit 120, but by increasing the filter frequency ω_f of the position-signal combination circuit 90, the utilization rate of the load position signal x_l is needed to be increased.

However, as represented in FIG. 2, because regarding the response G_l(s) of the load position signal x_l to the torque command signal τ_r, its phase delays more than that of the response G_p(s) of the motor position signal x_m to the torque command signal τ_r, regarding the open-loop transfer function L(s), its phase delays more than that of the semi-closed control system expressed by Equation 7. As a result, because the control system tends to be unstable, and its vibration also tends to increase, the filter frequency ω_f and the position gain k_p cannot have been sufficiently increased, Consequently, the stability compensation circuit 80 is configured in such a way that the compensated load-position signal x_lc in which compensation has been performed for the phase delay of the load position signal x_l is outputted.

FIG. 5 is a block diagram illustrating an internal configuration of the stability compensation circuit 80.

A second-order differential circuit 81 outputs the second-order differential signal of the load position signal x_l. A stability-compensation gain circuit 82 outputs a signal that is obtained by multiplying the output of the second-order differential circuit 81 by a stability compensation gain K_st set from the exterior. Moreover, the stability compensation circuit 80 outputs a summed signal, as the compensated load-position signal x_lc, obtained from the output of the stability-compensation gain circuit 82 and the load position signal x_l.

That is, the stability compensation circuit 80 performs calculation expressed by the following equation represented by a transfer function C_st(s).

$\begin{array}{cc}\left[\mathrm{Equation}9\right]& \\ \frac{x_{1c}}{x_1}=C_{\mathrm{st}}\left(s\right)=K_{\mathrm{st}}·s^2+1& \left(9\right)\end{array}$

Because the stability compensation circuit 80 operates as above, the transfer function from the torque command signal τ_r to the compensated load-position signal x_lc is expressed by the following equation.

$\begin{array}{cc}\left[\mathrm{Equation}10\right]& \\ \begin{array}{c}\frac{x_{1c}}{{\tau }_r}=\left(K_{\mathrm{st}}{s}^2+1\right)G_1\left(s\right)\\ =\frac{K_{\mathrm{st}}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\end{array}& \left(10\right)\end{array}$

Here, the stability compensation gain K_st is set by the following equation using the anti-resonance frequency ω_z of the control target 10.[Equation 11]K_st=1/ω_z²  (11)

Here, the anti-resonance frequency ω_z can be estimated by a method such as frequency-response measurement of the control target 10, or vibration-frequency measurement of the control target 10 when the speed gain k_v of the speed PI-control circuit 130 is increased.

When the stability compensation gain K_st is set as expressed by Equation 11, the transfer function from the torque command signal τ_r to the compensated load-position signal x_lc agrees with G_p(s) in Equation 4; moreover, the transfer function from the torque command signal τr to the control-target position signal xfb also agrees with G_p(s).

$\begin{array}{cc}\left[\mathrm{Equation}12\right]& \\ \frac{x_{\mathrm{fb}}}{{\tau }_r}=\frac{x_{1c}}{{\tau }_r}=\frac{x_m}{{\tau }_r}=G_p\left(s\right)& \left(12\right)\end{array}$

Therefore, the transfer function from the torque command signal τ_r to the basic control-torque signal τ_b agrees with the open-loop transfer function as expressed by the following equation, when the semi-closed control is performed as expressed by Equation 7.

$\begin{array}{cc}\left[\mathrm{Equation}13\right]& \\ \frac{{\tau }_b}{{\tau }_r}=-\frac{k_v\left(s+k_p\right)\left(s+{\omega }_{\mathrm{vi}}\right)}{s}\frac{{\omega }_s^{-2}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(13\right)\end{array}$

As a result of the above, stable control equivalent to the semi-closed control can be ensured by using the stability compensation circuit 80; moreover, the position gain k_p of the position gain circuit 120, and the filter frequency ω_f of the position-signal combination circuit 90 can be sufficiently increased.

As a result, control accuracy of the load position signal x_l can be improved.

Moreover, the anti-resonance points z₀ of the control target 10 expressed by Equation 6 are included in zeroes of the open-loop transfer function similarly to the case of the semi-closed control.

Here, a relationship between the stability compensation circuit 80 and high-frequency noises is explained.

The stability compensation circuit 80 is configured so as to output the compensated load-position signal x_lc by calculation including the second-order differential of the load position signal x_l. However, the control-target position signal x_fb is obtained through the load-position filter 92, and, because the response of the position control system is generally slower than that of the speed PI-control circuit 130, the filter frequency ω_f of the position-signal combination circuit 90 is needless to be extremely increased; therefore, the control-target position signal x_fb does not become extremely noisy.

The operation of the torque command signal τ_r in response to the input of the position command signal x_r is similar to the general semi-closed control; therefore, even if the position command signal x_r steeply varies, a problem never occurs in which the torque command signal τr steeply varies.

Moreover, by the damping compensation circuit 140, using the position command signal x_r, the motor position signal x_m, and the load position signal x_l, the damping-compensation torque signal τ_c calculated based on the damping adjustment parameter α set from the exterior is added to the basic control-torque signal τ_b. Thereby, the device is configured so as to obtain the torque command signal τ_r.

FIG. 6 is a block diagram illustrating an internal configuration of the damping compensation circuit 140.

A first damping gain circuit 141 receives as input a difference signal between the load position signal x_l and the motor position signal x_m, and outputs as a first damping compensation signal x_z1 a signal obtained by multiplying the difference signal by a first damping gain K_z1. A load-position differential circuit 142 outputs a signal obtained by differentiating the difference signal between the position command signal x_r and the load position signal x_l, while a second damping gain circuit 143 outputs as a second damping compensation signal x_z2 a signal obtained by multiplying the output from the load-position differential circuit 142 by a second damping gain K_z2. A third damping gain circuit 144 outputs as a third damping compensation signal x_z3 a signal obtained by multiplying the difference signal between the position command signal x_r and the load position signal x_l by a third damping gain K_z3. A damping adjustment circuit 145 outputs a signal obtained by multiplying by the damping adjustment parameter α a signal obtained by summing the first damping compensation signal x_z1, the second damping compensation signal x_z2, and the third damping compensation signal x_z3.

That is, the damping compensation circuit 140 performs the following calculation.[Equation 14]τ_c=α{K_z1(x_l−x_m)+K_z2·s·(x_r−x_l)+K_z3·(x_r−x_l)}  (14)

Next, a method of setting constant values for the damping compensation circuit 140 is explained. Due to the damping compensation circuit 140 that operates as described above, the transfer function from the torque command signal τ_r to the damping-compensation torque signal τ_c can be expressed by the following equation using Equation 6, Equation 7, and Equation 8.

$\begin{array}{cc}\left[\mathrm{Equation}15\right]& \\ \begin{array}{c}\frac{{\tau }_c}{{\tau }_r}=-\alpha \left\{K_{z1}\left(G_p\left(s\right)-G_1\left(s\right)\right)+\left(K_{z2}·s+K_{z3}\right)G_1\left(s\right)\right\}\\ =-\alpha \frac{K_{z1}{\omega }_z^{-2}{s}^2+K_{z2}s+K_{z3}}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\end{array}& \left(15\right)\end{array}$

The first damping gain K_z1, the second damping gain K_z2, and the third damping gain K_z3 of the damping compensation circuit 140 are set as follows using the speed gain kv, the speed-integration gain ωvi, and the position gain kp, which are constant values, set in the speed PI-control circuit 130 and the position gain circuit 120.[Equation 16]K_z1=k_vω_z²  (16)[Equation 17]K_z2=k_v(k_p+ω_vi)  (17)[Equation 18]K_z3=k_vk_pωvi  (18)

As a result of the above setting operation, the transfer function from the torque command signal τ_r to the damping-compensation torque signal τ_c can be expressed by the following equation.

$\begin{array}{cc}\left[\mathrm{Equation}19\right]& \\ \frac{{\tau }_c}{{\tau }_r}=-\frac{k_v\left\{s^2+\left(k_p+{\omega }_{\mathrm{vi}}\right)s+k_p{\omega }_{\mathrm{vi}}\right\}}{s}\frac{\alpha ·s}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(19\right)\end{array}$

Moreover, the open-loop transfer function obtained by opening the loop at the position of the torque command signal τ_r can be expressed by the following equation using Equation 15 and Equation 19.

$\begin{array}{cc}\left[\mathrm{Equation}20\right]& \\ \begin{array}{c}L\left(s\right)=\frac{{\tau }_b+{\tau }_c}{{\tau }_r}\\ =-\frac{k_v\left\{s^2+\left(k_p+{\omega }_{\mathrm{vi}}\right)s+k_p{\omega }_{\mathrm{vi}}\right\}}{s}\frac{{\omega }_z^{-2}{s}^2+\alpha ·s+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\\ =-\frac{k_v\left(s+k_p\right)\left(s+{\omega }_{\mathrm{vi}}\right)}{s}\frac{{\omega }_z^{-2}{s}^2+\alpha ·s+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}\end{array}& \left(20\right)\end{array}$

Therefore, the anti-resonance points of the open-loop transfer function vary from z₀ expressed by Equation 6 to z_c expressed by the following equation.[Equation 21]z_c=−ζ_zω_z±j·ω_z√{square root over (1−ζ_z²)}  (21)where a damping coefficient ζ_z at the anti-resonance points used in the above is expressed by the following equation.[Equation 22]ζ_z=α·ω_z/2  (22)

The pole and the zero point of the open-loop transfer function expressed by Equation 20 do not vary, according to varying of the damping adjustment parameter α, except for the anti-resonance points expressed by Equation 21 and equation 22. Moreover, the anti-resonance frequency ω_z as the absolute value of the anti-resonance points does not vary, but only the damping coefficient at the anti-resonance points varies.

By configuring the damping compensation circuit 140 as the above, only the damping coefficient at the anti-resonance points of the open-loop transfer function with respect to the control system in FIG. 1 is configured to be changed by the damping adjustment parameter α set from the exterior.

An advantage of this configuration is described as follows in which only the damping coefficient at the anti-resonance points of the open-loop transfer function is changed by the damping adjustment parameter α set from the exterior.

If the speed gain k_v is sufficiently increased, the closed-loop pole of the control system is known to asymptote to zero of the open-loop transfer function. That is, in a case in which the damping adjustment parameter α is set to zero in the semi-closed control system having the open-loop transfer function expressed by Equation 7, or in the control system in FIG. 1, when the speed control gain k_v is increased, a part of the closed-loop pole approaches the anti-resonance points of the control target 10 expressed by Equation 6. Accordingly, in the closed-loop pole, the damping coefficient becomes relatively low, and the response of the control target 10 becomes vibratory.

On the other hand, as expressed by Equation 22, when the damping adjustment parameter α is increased, the damping coefficient at the anti-resonance point increases. Moreover, the other zeroes of the open-loop transfer function L(s) are −ω_vi and −k_p being real numbers.

As a result, when the speed control gain k_v is increased, the closed-loop pole approaches the anti-resonance point where the damping coefficient is relatively large, and the real-number zero; therefore, even though an external disturbance is exerted to the control target 10, the vibration of the control system is controlled.

Here, the damping adjustment parameter α is set higher enough to such a degree that the damping coefficient ζ_z expressed by Equation 22 becomes approximately 0.5, so that the adjustment may be easy to be performed. Moreover, in order to increase the disturbance reduction effect, quite similar to the adjustment method in the general semi-closed control, the speed gain k_v, speed-integration gain ω_vi, and the position gain k_p may be increased.

Here, even if, in the damping compensation circuit 140, the first damping compensation signal x_z1 is calculated by multiplying a load acceleration signal, obtained by second-order differentiating the load position signal x_l, by the speed gain k_v and −1, because the open-loop transfer function agrees with Equation 20, similar effect can be obtained; however, because the second-order differential signal of the load position signal x_l is used, high-frequency noise components thereof increase. Therefore, as described above, by calculating the first damping compensation signal x_z1 by multiplying the difference signal between the motor position signal x_m and the load position signal x_l by the first damping gain K_z1, the device is configured so that the noise problems do not occur.

As described above, due to the effect of the stability compensation circuit 80 and the damping compensation circuit 140, not only by the adjustment performed by a method quite similar to that for the semi-closed control, but by a simple adjustment method in which the damping adjustment parameter α, as a parameter that is externally adjusted is only increased to an adequate value, the control accuracy of the load position signal x_l can be improved; consequently, a control system can be realized in which the vibration is reduced also against the external disturbance exerted to the control target 10.

Here, in the above description, the calculation of the block diagram represented in FIG. 6, that is, the calculation expressed by Equation 14 has been performed in the damping compensation circuit 140; however, similar effects can be obtained, even though the calculation expressed by the following equation for directly obtaining the damping compensation torque signal τ_c from the input signals into the damping compensation circuit 140.[Equation 23]τ_c=α(K_z2·s+K_z3)x_r−α(K_z2·s+K_z3−K_z1)x_l−α·K_z1·x_m  (23)

Moreover, the basis of the effect obtained by the damping compensation circuit 140 is that the vibration is reduced, by increasing the damping coefficient at the anti-resonance points of the open-loop transfer function, as expressed by Equation 22, so as to increase the damping coefficient of the closed-loop pole. Therefore, the device may be configured so that, if their open loop transfer functions are the same as each other, the calculation operation, in response to the position command signal x_r, in the damping compensation circuit 140 is different from those in Equation 14 and Equation 23.

For example, instead of the position command signal x_r inputted into the damping compensation circuit 140, a signal obtained by low-pass filtering the position command signal x_r may be used; thereby, varying of the damping-compensation torque signal τ_c corresponding to that of the position command signal x_r may be smoothed. In contrast, a command acceleration signal a_r as the second-order differential signal of the position command signal x_r is calculated inside the damping compensation circuit 140; then, a signal obtained by multiplying the command acceleration signal a_r by an adequate gain and the damping adjustment parameter α is further added to the damping-compensation torque signal τc. Thereby, the device may be configured so that the response of the motor position signal xm with respect to varying of the position command signal xr, when the damping adjustment parameter α is increased, becomes as fast as possible.

Embodiment 2

In Embodiment 1, the configuration for calculating the basic control-torque signal τ_b has been explained by Equations 1, 2, and 3 using the position gain circuit 120, the speed calculation circuit 100, and the speed PI-control circuit 130; however, the calculation may be performed using another configuration. Especially, when the transfer function from the torque command signal τr to the basic control-torque signal τ_b is equivalent to that expressed by Equation 13 in Embodiment 1, the calculation using the damping compensation circuit 140 may be performed as above without any modification.

On the other hand, when the transfer function from the torque command signal τ_r to the basic control-torque signal τ_b is different from that expressed by Equation 13, the calculation using the damping compensation circuit 140 may be correspondingly modified. The details are explained as follows.

For example, as measures to a case that the resolution of the motor position detector 60 is extremely low, assuming that the speed calculation circuit 100 calculates the motor speed signal v_m by calculation in which a speed filter F_v(s) is added as expressed by Equation 24 instead of the differential calculation of Equation 4, the transfer function from the torque command signal τ_r to the basic control-torque signal τ_b can be expressed by Equation 25.

$\begin{array}{cc}\left[\mathrm{Equation}24\right]& \\ v_m=s·F_v\left(s\right)·x_m& \left(24\right)\\ \left[\mathrm{Equation}25\right]& \\ \frac{{\tau }_b}{{\tau }_r}=-\frac{k_v\left\{\begin{array}{c}{s}^2{F}_v\left(s\right)+\\ \left(k_p+{\omega }_{\mathrm{vi}}{F}_v\left(s\right)\right)s+k_p{\omega }_{\mathrm{vi}}\end{array}\right\}}{s}\frac{{\omega }_z^{-2}{s}^2+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(25\right)\end{array}$

In this case, the calculation of the damping-compensation torque signal τ_c in the damping compensation circuit 140 may be performed by the following equation using the first damping gain K_z1 set by Equation 16, the third damping gain K_z3 set by Equation 18, and the damping adjustment parameter α, and using a fourth damping gain Kz4, a fifth damping gain Kz5, and a transfer function being the same as the speed filter F_v(s) in Equation 24, which are newly introduced.[Equation 26]τ_c=α{K_z1F_v(s)·(x_l−x_m)+(K_z4F_v(s)+K_z5)·s·(x_r−x_l)+K_z3·(x_r−x_l)}  (26)where the fourth damping gain K_z4 and the fifth damping gain K_z5 are respectively set as follows.[Equation 27]K_z4=k_vω_vi  (27)[Equation 28]K_z5=k_vk_p  (28)

Then, the open-loop transfer function L(s) obtained by opening at the position of the torque command signal τ_r can be expressed by the following equation.

$\begin{array}{cc}\left[\mathrm{Equation}29\right]& \\ L\left(s\right)=-\frac{k_v\left\{\begin{array}{c}{s}^2{F}_v\left(s\right)+\\ \left(k_p+{\omega }_{\mathrm{vi}}{F}_v\left(s\right)\right)s+k_p{\omega }_{\mathrm{vi}}\end{array}\right\}}{s}\frac{{\omega }_z^{-2}{s}^2+\alpha ·s+1}{J·s^2\left({\omega }_p^{-2}{s}^2+1\right)}& \left(29\right)\end{array}$

That is, by varying the damping adjustment parameter α, only the damping coefficient at the anti-resonance point among the zero points of the open-loop transfer function expressed by Equation 29 is configured to vary. Here, regarding the zero points except for the anti-resonance points, because if Equation 29 is expanded the description becomes complicated, the description is omitted; however, it is obvious that the coefficient does not vary even though the damping adjustment parameter α varies, and moreover, by a general adjustment, the zero points except for the anti-resonance points become real numbers, or complex zeroes whose damping coefficient is relatively large.

Accordingly, even when the transfer function from the torque command signal τ_r to the basic control-torque signal τ_b is different from that expressed by Equation 13 in Embodiment 1, if the speed gain k_v is sufficiently large, only by increasing the damping adjustment parameter α to an adequate value similarly to that in Embodiment 1, the damping coefficient of the closed-loop pole can be increased. Thereby, even when an external disturbance is exerted to the control target 10, the vibration can be reduced; consequently, the position of the motor 30 or the position of the load 20 can be accurately controlled by a simple adjustment operation.

Embodiment 3

In Embodiment 1, by summing the load position signal x_l and the multiplied signal obtained by multiplying by the stability compensation gain K_st the signal obtained by second-order differentiating the load position signal x_l, the compensation has been performed for the phase delay of the load position signal x_l; however, if a method in which another calculation is performed also has the effect of advancing the phase at a frequency close to the anti-resonance frequency ω_z and the resonance frequency τ_p of the control target 10, even though the effect is large or small, the compensation can also be performed thereby for the phase delay of the load position signal x_l.

FIG. 7 is a block diagram illustrating a machine position control device according to Embodiment 3 of the present invention, in which the damping compensation circuit 140 of Embodiment 1 represented in FIG. 1 is omitted so that the basic control-torque signal τ_b is configured to be used as the torque command signal τ_r without any modification.

For example, in a stability compensation circuit 80a, calculation expressed by the following transfer function is performed using the stability compensation gain K_st and a filter time-constant t_st.

$\begin{array}{cc}\left[\mathrm{Equation}30\right]& \\ x_{1c}=\frac{K_{\mathrm{st}}{s}^2+1}{{\left(t_{\mathrm{st}}s+1\right)}^2}·x_1& \left(30\right)\end{array}$

Equation 30 is a calculation equation in which a secondary low-pass filter is further added to the stability compensation circuit 80 in Embodiment 1, by which more effect can be obtained for reducing noises when any problem occurs caused by especially low resolution of the load position detector 70.

In this case, the stability compensation gain K_st may be set to a value close to that in Equation 11 used in Embodiment 1; moreover, if the filter time-constant t_st is set to a value smaller than the root of the stability compensation gain K_st, an effect of advancing the phase can be obtained by the calculation of Equation 30.

Here, in Equation 30, the calculation has been performed that includes phase-advancing characteristics due to the denominator and the numerator each being a secondary transfer function; however, even though the denominator and the numerator each being a primary transfer function, for example, as expressed by the following equation, because an effect of advancing the phase at a frequency close to the anti-resonance frequency ω_z and the resonance frequency ω_p can be obtained, an effect that the position gain k_p and the filter frequency ω_f can be increased compared to a configuration in which the stability compensation circuit 80a is not provided can be obtained.

$\begin{array}{cc}\left[\mathrm{Equation}31\right]& \\ x_{1c}=\frac{t_1·s+1}{t_2·s+1}·x_1& \left(31\right)\end{array}$

Here, the effect of advancing the phase can be obtained when t₂<t₁; moreover, t₁ is assumed to be set to a value close to the inverse number of the anti-resonance frequency ω_z.

By configuring as described above, the stability compensation circuit 80a receives as input the load position signal x_l, and outputs, based on the calculation of the phase-advancing transfer function expressed by Equation 30 or Equation 31, the compensated load-position signal xlc. Moreover, the position-signal combination circuit 90 outputs as the control-target position signal x_fb a summed signal obtained by summing a signal obtained by high-pass filtering the motor position signal x_m and a signal obtained by low-pass filtering the compensated load-position signal xlc. Therefore, in the configuration according to Embodiment 3, the calculation with respect to the torque command signal τ_r is performed based on the position command signal x_r and the control-target position signal x_fb.

As a result, the control-target position signal x_fb whose phase is more advanced compared to that of a configuration provided without the stability compensation circuit 80a is fed back thereto, and, even if the mechanical rigidity of the control target 10 is relatively low, the filter frequency ω_f of the position-signal combination circuit 90 and the position gain k_p of the position gain circuit 120 can be stably increased; consequently, the control accuracy of the load position signal x_l can be improved.

Embodiment 4

An operation of the damping compensation circuit 140 when the stability compensation circuit 80 and the position-signal combination circuit 90 in Embodiment 1 are omitted is as follows.

FIG. 8 is a block diagram illustrating a machine position control device according to Embodiment 4, in which the stability compensation circuit 80 and the position-signal combination circuit 90 are omitted from the configuration represented in FIG. 1, and the motor position signal x_m instead of the control-target position signal x_fb is directly inputted into the position gain circuit 120.

In the above configuration, the open-loop transfer function can be expressed by Equation 20 similarly to that in Embodiment 1, and the device is configured in such a way that only the damping coefficient ζ_z at the anti-resonance points of the open-loop transfer function is varied by varying the damping adjustment parameter α. As a result, if the speed gain k_v is sufficiently large, only by adequately increasing the damping adjustment parameter α, the damping coefficient of the closed-loop pole can be increased; consequently, even in a case in which an external disturbance is exerted to the control target 10, the vibration can be reduced.

Here, in Embodiment 4, because the stability compensation circuit 80 and the position-signal combination circuit 90 are omitted, a function for operating in such a way that the motor position signal x_m agrees with the position command signal x_r, and compensates the steady error occurring between the load position signal x_l and the motor position signal x_m is not provided; however, in application in which such error is not treated as a problem, the vibration can be reduced by increasing the speed gain k_v and the position gain k_p using a simple configuration of the control system, and by increasing the damping coefficient at the anti-resonance points of the open-loop transfer function using the function of the damping compensation circuit 140; consequently, the position of the motor 30 can be accurately controlled. Moreover, as a result, the position of the load 20 can also be controlled without problems. Therefore, the position of the motor 30 and the load 20 can be accurately controlled by a simple adjustment operation.

Here, except for configurations represented in Embodiment 1 through Embodiment 4, modified examples such as a configuration that the motor speed signal v_m as speed feedback of the motor is not inputted into the position-speed control circuit 110, and a configuration that a circuit for performing speed IP control is used instead of the speed PI-control circuit 130, etc. can be proposed; thereby, the stability compensation circuit 80 and the damping compensation circuit 140 having transfer characteristics corresponding to the changes can be configured similarly to those in Embodiment 1 through Embodiment 4.

Claims

1. A machine position control device comprising: a position-speed control means for receiving as input a position command signal indicating a position target value of a load driven by a motor, and a control-target position signal as reference information related to present positions of the motor and the load, and for outputting a torque command signal as a target value of torque at which the motor drives the load;a stability compensation means for receiving as input a load position signal as a present-position measurement value of the load, and for outputting a compensated load-position signal after a phase delay of the load position signal has been compensated based on a phase-advance transfer function; anda position-signal combination means for combining a signal including a high-frequency component of a motor position signal as a present-position measurement value of the motor, with a signal including a low-frequency component of the compensated load-position signal, and for outputting the control-target position signal.

2. A machine position control device as recited in claim 1, wherein the stability compensation means sums the load position signal and a signal obtained by multiplying by a stability compensation gain a signal having been obtained by second-order differentiating the load position signal.

3. A machine position control device as recited in claim 1, wherein the position-speed control means also receives as input a motor speed signal indicating a present speed of the motor.

4. A machine position control device comprising: a position-speed control means for receiving as input a position command signal indicating a position target value of a load driven by a motor, and a control-target position signal as reference information related to present positions of the motor and the load, and for outputting a basic control-torque signal to be a basis for calculating a torque command signal as a target value of torque at which the motor drives the load;a stability compensation means for receiving as input a load position signal as a present-position measurement value of the load, and for outputting a compensated load-position signal after a phase delay of the load position signal has been compensated based on a phase-advance transfer function;a position-signal combination means for combining a signal including a high-frequency component of a motor position signal as a present-position measurement value of the motor, with a signal including a low-frequency component of the compensated load-position signal, and for outputting the control-target position signal; anda damping compensation means, being means for receiving as input the motor position signal and the load position signal and for outputting a damping-compensation torque signal to which the basic control-torque signal is added for obtaining the torque command signal, for calculating the damping-compensation torque signal, when a load drive system with the motor is modeled as a two-inertia system, based on a second open-loop transfer function determined, in response to a first open-loop transfer function as a transfer function from the torque command signal to the basic control-torque signal, so that an open-loop transfer function obtained by combining with the first open-loop transfer function agrees with a function in which only a damping coefficient at an anti-resonance point of the two-inertia system of the first open-loop transfer function is varied, so that the transfer function from the torque command signal agrees with the second open-loop transfer function.

5. A machine position control device as recited in claim 4, wherein the damping-compensation torque signal outputted from the damping compensation means is a signal obtained by multiplying, by a coefficient based on a damping adjustment parameter for adjusting the damping coefficient at the anti-resonance point of the two-inertia system, a signal obtained by summing a signal obtained by multiplying by a first damping gain the difference signal between the load position signal and the motor position signal, a signal obtained by multiplying by a second damping gain a signal obtained by differentiating the difference signal between the position command signal and the load position signal, and a signal obtained by multiplying by a third damping gain the difference signal between the position command signal and the load position signal.

6. A machine position control device comprising: a position-speed control means for receiving as input a position command signal indicating a position target value of a load driven by a motor, and a motor position signal as a present-position measurement value of the motor, and for outputting a basic control-torque signal to be a basis for calculating a torque command signal as a target value of a torque when the motor drives the load; anda damping compensation means, being means for receiving as input the motor position signal and a load position signal as a present-position measurement value of the load, and for outputting a damping-compensation torque signal to which the basic control-torque signal is added for obtaining the torque command signal, for calculating the damping-compensation torque signal, when a load drive system with the motor is modeled as a two-inertia system, based on a second open-loop transfer function determined, in response to a first open-loop transfer function as a transfer function from the torque command signal to the basic control-torque signal, so that an open-loop transfer function obtained by combining with the first open-loop transfer function agrees with a function in which only a damping coefficient at an anti-resonance point of the two-inertia system of the first open-loop transfer function is varied, so that the transfer function from the torque command signal agrees with the second open-loop transfer function.