Information
-
Patent Grant
-
4950967
-
Patent Number
4,950,967
-
Date Filed
Monday, April 3, 198935 years ago
-
Date Issued
Tuesday, August 21, 199034 years ago
-
Inventors
-
Original Assignees
-
Examiners
- Roskoski; Bernard
- Bergmann; Saul M.
Agents
-
CPC
-
US Classifications
Field of Search
US
- 318 567
- 318 561
- 318 615
- 318 616
-
International Classifications
-
Abstract
The invention provides a servomotor control apparatus which performs standard mode-type velocity control. An acceleration signal is formed by a differential computation or difference computation, from a velocity signal fed back from a servomotor. An acceleration value of the servomotor (d) is estimated by a criterion model (2) of a velocity control system utilizing a current command T(S) formed in a velocity control loop. The computed acceleration and the estimated acceleration are compared and the current command T(S) is corrected in conformity with the results of the comparison.
Description
BACKGROUND OF THE INVENTION
This invention relates to a servomotor control apparatus having improved characteristics, such as a frequency response characteristic.
DESCRIPTION OF THE RELATED ART
In servomotor velocity control, ordinarily it is so arranged that a velocity signal fed back from the servomotor is compared with a velocity command. FIG. 4 is a block diagram of an example of such a conventional control system in a case where disturbances are neglected. In FIG. 4, A represents an integration term having an integration gain K.sub.1, B represents a proportion term having a proportional gain K.sub.2, C represents a current amplification term of torque constant K.sub.T, and D represents a motor term which includes motor rotor inertia J.sub.M and a constant .beta. indicating load inertia. Further, S represents an integration operator.
With this velocity control system, the sum of the result of integrating an error signal between a velocity command V(s) and motor velocity W(s) by K.sub.1 /S of the integration term and the result of feeding back the motor velocity W(s) through the proportional gain K.sub.2 of the proportion term B is applied to the motor D as a torque command T(s) (current command).
In such a velocity control system, a time delay develops in the response of the torque command T(s) owing to the integration term A, and instability occurs such as control system oscillation when the system gain is enlarged. More specifically, the system cannot quickly follow up rapid changes in the commanded velocity V(s), and oscillation is produced particularly when the motor is controlled to be brought to a stop.
SUMMARY OF THE INVENTION
The present invention has been devised in order to solve the foregoing problems and its object is to provide a servomotor control apparatus in which frequency response is raised and velocity can be stably controlled by mitigating the effects of disturbances and fluctuations in the load.
In accordance with the invention, there is provided a servomotor control apparatus having a velocity control loop which includes a control criterion model for performing standard model-type control with regard to a servomotor. The servomotor control apparatus includes means for forming an acceleration signal by differentiating a velocity signal fed back from the motor, means for estimating an acceleration value of the servomotor using a current command output by the velocity control loop, means for obtaining an error between the acceleration signal and the estimated acceleration value, and means for correcting the current command by the error signal.
Accordingly, the servomotor control apparatus of the present invention compares a computed acceleration and the estimated acceleration and enables the current command to be corrected in conformity with the results of the comparison to stabilize velocity control.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of an embodiment of the invention having a standard model-type velocity control loop;
FIG. 2 is a block diagram of a velocity control system which takes the effects of disturbances into account;
FIG. 3 is a block diagram of a parallel-type model standard control;
FIG. 4 is a block diagram of a conventional velocity control loop.
DESCRIPTION OF THE PREFERRED EMBODIMENT
An embodiment of the present invention will now be described in detail with reference to the drawings.
Since model standard-type adaptive control is used in the present invention, the principle of control will be described first.
FIG. 3 is a block diagram of an example of parallel-type model standard control. As shown in FIG. 3, a criterion model 2 is arranged in parallel with a controlled system 1. An adjustable gain 3 provided ahead of the controlled system 1 is adaptively corrected during operation of the controlled system 1. Gain control is performed by an adaptive controller 4 in conformity with the output of an arithmetic unit 5 in such a manner that an output I.sub.OUT of the controlled system 1 will follow up an output I.sub.OUT of the criterion model 2. I.sub.REF represents a command signal, and E denotes an error signal. Such model standard-type adaptive control is advantageous in that adaptive control of the control system can be performed with a quick response.
FIG. 1 is a block diagram of an embodiment of the present invention. Portions identical with those of FIGS. 3 and 4 are designated by like reference characters and a description thereof is deleted: This embodiment of the invention has a model standard-type velocity control loop and is characterized in that a coefficient .alpha. of the criterion model 2 of the velocity control system is set to a value at which the cut-off frequency and damping coefficient of the velocity control loop will be made equal to what they are when a corrective gain coefficient K.sub.A =0 is maintained. The corrective gain K.sub.A is provided as the adjustable gain 3. That is, along with the velocity feedback loop, an acceleration feedback loop is formed by providing an arithmetic unit 6 having the differentiation operator s to differentiate the motor velocity W(s). Acceleration is estimated by the criterion model 2 based on the current command T(s). The acceleration is compared with the fed back acceleration signal by the arithmetic unit 5 to correct the current command T(s).
Described next will be a transfer function H(s)= W(s)/V(s) in the control loop of FIG. 1.
In FIG. 1, the relationship among a velocity command Vcmd (=V(s)), torque command T(s), corrective torque signal A(s) and motor velocity W(s) is as follows:
T(s)=(K.sub.1 /S)V(s)-{(K.sub.1 /S)+K.sub.2 }W(s)
A(s)=K.sub.A {(.alpha.K.sub.T /J.sub.M)T(s)-SW(s)}
W(s)=(K.sub.T {SJ.sub.M (1+.beta.)}){T(s)+A(s)}
The transfer function H(s) is as follows: ##EQU1##
(A) With respect to the transfer function H.sub.0 (s) when the corrective gain K.sub.A =0 this corresponds to the conventional example shown in FIG. 4, that is, an arrangement in which there is no acceleration feedback system. Assuming that the transfer function H.sub.0 (s)=W(s)/V(s), we have
H0(s)=(K.sub.T K.sub.1)/{J.sub.M (1+.beta.)S.sup.2 +K.sub.T K.sub.2 S+K.sub.T K.sub.1 } (2)
Here the integration term A and proportion term B are adjusted in conformity with the load inertia ratio, so that we have
K.sub.1 =(1+.beta..sub.0)K.sub.10, K.sub.2 =(1+.beta..sub.0)K.sub.20
where .beta..sub.0 : constant corresponding to the load inertia ratio
If the constant .beta. indicating the load inertia is as follows:
.beta.=.beta..sub.0
Eq. (2) can be expressed as follows: ##EQU2##
(B) With respect to the transfer function H(s), when the corrective gain K.sub.A .noteq.0 and the coefficient .alpha.=0, here system operation will be described when the conventional velocity feedback system is provided solely with the acceleration feedback function shown in FIG. 1. Assuming that the transfer function is H(s)=W(s)/V(s), we have
H(s)=(K.sub.T K.sub.1)({J.sub.M (1+.beta.)+K.sub.T K.sub.A }S.sup.2 +K.sub.T K.sub.2 S+K.sub.T K.sub.1 (4)
This indicates that, in comparison with Eq. (2), the provision solely of acceleration feedback is equivalent to effecting a pseudo-increase in motor rotor inertia.
(C) With respect to the transfer function H(s), when the corrective gain K.sub.A .noteq.0 and the coefficient .alpha..noteq.0.
This corresponds to a case where model standard control shown in FIG. 1 is provided with acceleration feedback. By making the criterion mode coefficient .alpha. equal to 1/(1+.beta.), the behavior exhibited is the same as that in the conventional velocity control system with respect to the velocity command, and the apparent motor rotor inertia is enlarged so that the system presents improved stability. That is, the transfer function H(s) becomes as follows, where the substitution (.alpha.K.sub.T /J.sub.M)=.alpha.' is made in Eq. (1): ##EQU3##
Next, when K.sub.A is decided in such a manner that K.sub.T K.sub.A /J.sub.M (1+.beta.)=P holds, namely when the apparent rotor inertia becomes (1+P) times larger, .alpha. becomes as follows from K.sub.a .alpha.'=P
.alpha.=1/(1+.beta.)
Accordingly, the transfer function H(s) is ##EQU4##
More specifically, by setting .alpha. to be equal to 1/(1+.beta.), Eq. (5) is such that the velocity loop exhibits the same behavior as when K.sub.A =0 (the prior-art example) with respect to the velocity command V(s), and the apparent motor rotor inertia becomes larger at the same time so that the velocity loop exhibits improved velocity control system stability.
FIG. 2 is a block diagram of a velocity control system that takes the influence of disturbance into account. Before investigating the effects of disturbance, the influence of a change in inertia on the transfer function will be considered.
Described first will be the influence of a fluctuation in inertia on the transfer function H(s) when the motor inertia coefficient .beta. is represented by .beta.(s). The transfer coefficient H(s) may be written as follows from Eq. (1): ##EQU5## The influence of the fluctuation in inertia on the transfer function H(s) is evaluated by the following equation: ##EQU6## If K.sub.A =0 holds, we have the following from Eq. (2): ##EQU7##
Next, the following is obtained from Eqs. (6), (7):
G.sub.I (s)={.differential.H(s)/.differential..beta.(s)}/{.differential.Ho(s)/.differential..beta.(s)} (8)
If the coefficient .alpha. of the criterion model is made equal to 1/(1+.beta..sub.1), as mentioned earlier, then we will have the following since K.sub.A =P(1+.beta..sub.1).times.(J.sub.M /K.sub.T) ##EQU8## Though .beta.(s) fluctuates in the vicinity of .beta..sub.1, we arrive at the following if the range of fluctuation is neglected and 1+.beta.(s)=1+.beta..sub.1 is assumed to hold: ##EQU9## Eq. (9) indicates that when the corrective gain K.sub.A is determined in such a manner that the motor rotor inertia becomes (1+P) times larger and .alpha. is determined in such a manner that the cut-off frequency and damping coefficient of the velocity control system become equal to what they will be at K.sub.A =0, the influence of the change in load inertia upon the change in the velocity loop transfer function can be reduced to a value which is 1/(1+P) times as much.
The influence of a variation in torque upon output velocity will now be considered on the basis of a transfer function for a case where a torque disturbance T.sub.D (s) is taken into account in FIG. 2.
If it is assumed that a torque command T(s)=(K.sub.1 /S)V(s)-{(K.sub.1 /S)+K.sub.2 }W(s), a torque command signal A(s)=K.sub.A {(.alpha.K.sub.T /J.sub.M)T(s)-SW(s)} and a motor velocity W(s)=(1/{SJ.sub.M (1+.beta.)})(T.sub.D (s)+K.sub.T {T(s)+A(s)}), then the following equation will hold with regard to motor velocity W(s): ##EQU10##
Solving Eq. (10) with regard to motor velocity W(s), we have ##EQU11## The manner in which the motor velocity W(s) varies due to a change in torque disturbance T.sub.D (s) will now be investigated based on the foregoing. Specifically, we have ##EQU12## Letting W.sub.0 (s) represent motor velocity when K.sub.A =0 prevails, we have ##EQU13## Accordingly, we have the following from Eqs. (12), (14): ##EQU14##
Accordingly, if the coefficient .alpha. of the criterion model is made equal to 1/(1+.beta.) and K.sub.A =P(1+.beta.)(J.sub.M /K.sub.T) holds, then we have ##EQU15## Eq. (9) indicates that when the corrective gain K.sub.A is determined in such a manner that the motor rotor inertia becomes (1+P) times larger and .alpha. is determined in such a manner that the cut-off frequency and damping coefficient of the velocity control system become equal to what they are when K.sub.A =0 holds, the influence of the change in load inertia upon the change in the velocity loop transfer function can be reduced to a value which is 1/(1+P) times as much.
Though an embodiment of the invention has been described, the invention is not limited thereto but can be modified in various ways without departing from the scope of the claims.
The servomotor control apparatus of the present invention is such that the velocity control system of a servomotor is provided with a loop which forms an acceleration signal by differentiating motor velocity, acceleration serving as a model criterion is specified using a current command, this estimated value and an acceleration signal are compared and the current command is corrected. As a result, frequency response is improved and the influence of disturbance and torque variation can be diminished to make stable control possible.
Claims
- 1. A servomotor control apparatus having a velocity control loop which includes a control criterion model for performing standard model-type control with regard to a servomotor, comprising:
- means for forming an acceleration signal by differentiating a velocity signal fed back from said motor;
- means for estimating an acceleration value of the servomotor using a current command output by the velocity control loop;
- means for obtaining an error between said acceleration signal and the estimated acceleration value and outputting an error signal; and
- means for correcting said current command in accordance with said error signal.
- 2. A servomotor control apparatus according to claim 1, wherein a corrective gain is provided with respect to the error between said acceleration signal and the estimated acceleration value, said error signal being obtained by setting a corrective gain coefficient K.sub.A in conformity with a fluctuation in load inertia.
- 3. A servomotor control apparatus according to claim 2 wherein said means for estimating the acceleration value of said servomotor employs a control criterion model of .alpha. (K.sub.T /J.sub.M, where K.sub.T represents a torque constant and J.sub.M represents motor rotor inertia, the coefficient .alpha. being set to a value at which a cut-off frequency and damping coefficient of the velocity control loop will be made equal to what they are when the corrective gain coefficient is K.sub.A =0.
Priority Claims (1)
Number |
Date |
Country |
Kind |
62-217344 |
Aug 1987 |
JPX |
|
PCT Information
Filing Document |
Filing Date |
Country |
Kind |
102e Date |
371c Date |
PCT/JP88/00867 |
8/31/1988 |
|
|
4/3/1989 |
4/3/1989 |
Publishing Document |
Publishing Date |
Country |
Kind |
WO89/02185 |
3/9/1989 |
|
|
US Referenced Citations (4)
Number |
Name |
Date |
Kind |
4680518 |
Kurakake et al. |
Jul 1987 |
|
4695780 |
Kurakake et al. |
Sep 1987 |
|
4816734 |
Kurakake et al. |
Mar 1989 |
|
4851754 |
Sakamoto et al. |
Jul 1989 |
|