JP2824576B2 - Servo control device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a servo control device, and more particularly, to a servo control device that performs feedforward control by increasing the accuracy of estimating the amount of disturbance applied to a control target. [Prior Art] As a conventional servo control device, disclosed in Japanese Patent Application Laid-Open No. 60-153503, an amount of disturbance added to a control target is estimated and an internal state quantity to be present corresponding to the amount of disturbance is calculated. There is known a method of performing optimal feedback control by using a difference between a calculated internal state quantity and an actual internal state quantity. Further, the amount of disturbance applied to the motor shaft, which varies according to the rotation angle position of the motor shaft described in JP-A-62-28803, is stored in a memory of a computer, and the rotation of the motor shaft is controlled. There is also known a device that takes out the amount of disturbance in the memory in accordance with the position angle and performs feedforward control to prevent the disturbance from affecting the electric motor. [Problems to be Solved by the Invention] The method described in Japanese Patent Application Laid-Open No. 60-153503 always seeks the movement of the equilibrium point of the control system based on disturbance, and shifts the origin of the equilibrium point. A new target has been set, which is
Formula (19) described in JP-A-60-153503, It is clear from. Internal state at r equilibrium in the above equation is (target value), u _S operation input value that resulted in the equilibrium state (operation amount target value), x _S is the internal state of the target value to be also referred to as the equilibrium point , W are disturbances, and A, B, C and E are coefficients. That is, the balance of the new target value x _S and the disturbance w, and calculates the operation input value u _S to the state quantity x to zero. In the servo system, even if the state quantity x attempts to become the command value _xREF , x is set to 0 by the feedback control, so that the followability to the command value deteriorates. Further, the apparatus described in Japanese Patent Application Laid-Open No. 62-28803 uses a disturbance amount corresponding to the angular position of the rotating shaft of the electric motor, and cannot respond to disturbance fluctuation due to angular velocity, angular acceleration, etc., and estimates the disturbance amount. Is difficult. SUMMARY OF THE INVENTION An object of the present invention is to provide a servo control device capable of improving the followability to a change in a state quantity command value. [Means for Solving the Problems] In order to achieve the above object, the present invention provides a state variable for a controlled object having a plurality of internal state variables, wherein the internal state variable changes with a change in a state variable command value. A steady state deviation compensating means for generating a control output by inputting a deviation between a command value and a state quantity actual value indicating an actual state quantity of the controlled object; and estimating an internal state quantity command value based on the state quantity command value. Command value estimating means, an internal state quantity deviation calculating means for calculating an internal state quantity deviation between the estimated internal state quantity command value and an internal state quantity by the output of the controlled object, and the control output and the internal state quantity And a control operation means for controlling a control target based on the deviation. [Operation] According to the means described above, when the state quantity command value for the major loop is in the steady state, the internal state quantity of the control target is also held in the steady state, and the state quantity actual value indicating the actual state quantity of the control target is A control output based on the deviation from the state quantity command value is generated, and the control target is controlled based on the control output. On the other hand, when the state quantity command value for the major loop changes, the internal state quantity of the control target also changes in response to this change. When the change in the internal state quantity is fed back via the minor loop as it is, the change in the internal state quantity acts as a contradictory value to the control output based on the change in the state quantity command value. That is, if the change in the internal state quantity is fed back via the minor loop as it is, the change in the control output is suppressed by the change in the internal state quantity, and it becomes impossible to immediately respond to the change in the state command value. However, when the state quantity command value changes, the internal state quantity command value corresponding to the change in the state quantity command value is estimated based on the state quantity command value, and the estimated internal state quantity command value is used as the control target. When a control is added to the internal state quantity to obtain a deviation between the estimated internal state quantity command value and the internal state quantity of the control target, and the deviation is used as a feedback amount of the minor loop, the estimated internal state quantity command is obtained. The command value (control output) for the minor loop is corrected according to the deviation between the value and the internal state quantity of the control target. That is, the estimated internal state quantity command value functions to suppress a change in the internal state quantity relating to the control target, and the deviation between the two is obtained as a value approaching zero. For this reason, even if the state quantity command value for the major loop is changed, it is possible to prevent the command value (control output) for the minor loop and the value of the feedback amount in the minor loop from becoming contradictory values.
Therefore, even if the state quantity command value changes, the ability to follow the change in the state quantity command value can be improved. Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a system diagram showing a basic configuration of the present invention. In FIG. 1, a disturbance d (vector, hereinafter the same) and a control output u (vector, hereinafter the same) are added to the controlled object 1, and a state quantity x (vector, hereinafter the same) is output from the controlled object 1. You. The state quantity x is further multiplied by the matrix C in the arithmetic unit 2 to obtain a state quantity y with respect to the command value y _REF applied to the servo system. A deviation e = y _REF −y between the state quantity y and the command value y _REF and the steady-state deviation compensator 3 constitute a servo system. Since the disturbance d is applied to the control target 1 as described above, the state quantities x and y fluctuate. An observer 4 is provided which receives a control output u applied to the control target and a state quantity x of the control target and outputs an estimated value of the state quantity. If all the state quantities x can be detected, no observer is provided and the detected value is used. The disturbance d is expressed in the form of d = f (x) based on a model quantifying the dynamic characteristics of the apparatus, and a disturbance estimation provided with a state quantity including the estimated value or a state quantity including only the detected value as an input. The calculator 5 outputs the estimated disturbance by performing the calculation of = f (x), and the calculator 10 multiplies the feedforward control gain M (matrix, the same applies hereinafter) by M ·
Is output and added to the control output, so that feedforward control is performed on the controlled object by estimating the disturbance, and the influence of the disturbance on the controlled object is reduced. In the command value estimator 7,
Y = y using the control target model with the command value y _REF as input
The control output u ₀ and the internal state quantity x _{0 in} the steady state at _REF
Look, adding a control output u ₀ to the control output. At the same time, the deviation calculator 11, the internal state quantity x ₀ obtained by the command value estimator 7 from the internal state quantity inputted to the disturbance estimator 5 subtracted, multiplied by the feedback control gain to the deviation obtained by the feedback A control output is used, and feedback control is performed by adding the control output to the control internal force. If the feedback control output is applied to the control output, a feedback control system controls attempts x to x _0, thus followability is improved with respect to the command value y _REF. Hereinafter, an embodiment in which the present invention is applied to a swing motor control of a sizing press in a steel plant will be described. As shown in FIG. 2, the sizing press is an apparatus that presses a leading end of a material to be pressed 207, which is a rolled material, in a traveling direction using a mold 202 to facilitate rolling. The material to be pressed 207 is
Mold 202 driven by main motor 201 in the direction of arrow 205
However, since the material to be pressed 207 is moving in the direction of the arrow 208, the die 202 is pressed while the material to be pressed 2 is pressed.
It must be moved in the same direction and at the same speed as 07 in the swing direction indicated by arrow 206. Therefore, a swing motor 203 that swings the mold 202 in the swing direction 206 and a crank supported by the fulcrum 204 are provided. If the rotation speed of the swing motor 203 is the same, the swing direction of the mold 202 during pressing is provided. The speed to 206 is fixed. Therefore, since it is desired that the oscillating motor 203 rotates at a constant speed, the load torque applied to the oscillating motor 203 fluctuates due to the fluctuation of the moment due to a mechanical system such as a crank or a press, and this torque fluctuation becomes a disturbance. Swinging motor 203
In order to prevent the rotation speed from deviating from the command value (constant speed), a control device to which the present invention is applied as shown in FIGS. 3 and 4 is provided. FIG. 3 shows a block diagram of a control target of the oscillating motor.
Mechanical system phenomena such as cranks and presses
While generating a disturbance T _S for the rotational speed of the T _S is theta ,, and related functions (theta is swinging motor 203,
Is the rotational angular velocity and is the rotational angular acceleration).
The state equation of the motor control system and the control target of the motor system is expressed by the following equation. The disturbance T _S is estimated by a disturbance estimator. The state equation of the controlled object is Next, if the state quantity x = [theta ,,] ^{T, = Ax + Bu + ET} S ......... (2) Toke, has a controllable and observable. A, B, and E
Are coefficients, each of which is a matrix. Generally, it is necessary to operate the state quantity y, that is, x, so as to follow the command value by the command value y _REF of the servo system.
(From y _REF = Cx _REF , x _REF = C ^-1 y _REF must be satisfied.) However, when feedback control is performed on the control target, the control system tries to set the state quantity x to O, and x _REF
Will not follow. Therefore, the control output u ₀ and the state quantity x ₀ in the steady state where y = y _REF is _obtained from the command value y _fRE using the control object model, and u ₀ is added to the control output, and x ₀ is estimated as the state quantity. Subtract feedback control from the value.
When the feedback control output and u _{F · B,} u _FB is, u
_FB = represented by F (x-x _0). If the u _FB is applied to the control output, a feedback control system controls attempts x to x _0, thus followability is improved with respect to the command value y _REF. The control system is considered as a servo system related to θ, and y = θ = [100] x = Cx (3) In order to make the servo problem a regulator problem, a variable Z is newly provided, and = θ _REF −y = θ _REF −θ (4) θ _REF is a command value. From equations (1) and (4), the state equation is: Becomes This equation of state can be expressed as
The optimal feedback gain F = [F ₁ , F ₂ , F ₃ , F ₄ ] and the feedforward gain M ₁ for the disturbance T _S are determined by solving the i-equation using a method or the like. The disturbance T _S is an equation modeling the mechanical physics of FIG. Required by However, v _X = {r ₁ sin θ + 2r ₂ cos (2θ)} (v _X is a mold 202
Rocking speed) alpha _X = a _{{1 r cosθ-4r 2 sin} (2θ)} + rocking acceleration of _{{r 1 sinθ + 2r 2 cos} (2θ} (α X mold 202) g: gravitational acceleration W _S: Gold Weight of the mold 202 GD ² : Moment of inertia F ₁ : Press force F ₂ : Friction force Since only θ can be detected in this controlled object, an observer 4 having u, θ, and as inputs is provided. hand Confirm. Various literatures have been published on the method of forming the observer, and a description thereof will be omitted. In the steady state, since θ = θ _REF , if the disturbance T _S can be completely controlled by the feedforward control and is ignored, the state equation (5) becomes Become. Here, θ ₀ , ₀ , ₀ , u ₀ are the values of the state quantities θ,, and the control output u when the steady state is established with respect to the command value θ _REF , and these θ ₀ , ₀ , ₀ , u ₀ is based on equation (7), Required by The determined u ₀ is added to the control output u, and the determined θ ₀ , ₀ , and ₀ are
From the actual value θ and the estimated value And the optimum feedback gain for this difference
F ₁ , F ₂ and F ₃ are multiplied and added to the control output. Disturbance T
_S is θ, and Is input based on the disturbance estimator 5 and is output as an estimated disturbance T _S based on the equation (6). The input disturbance E is multiplied by the feedforward gain M ₁ by the calculator 10 and supplied to the control output. In FIG. 4, the measurable θ and the control output u among the θ added to the mechanical system physical phenomenon 8 from the controlled object 1 and the control output u are input to the observer 4. Is output. In addition, the above measurable θ, and Is input to the disturbance estimator 5, and the estimated disturbance T _S is output based on the above equation (6). Further, the command value estimator 7 gives the command value θ
_REF is input, the (8) control output u ₀ in the steady state with respect to the command value comprising theta _REF based on the type, state quantity theta _{0, 0,} and ₀ is outputted. Said u ₀ is added to the control output u,
θ ₀ , ₀ , and ₀ are the measured θ, ₀ , and the estimated value by the deviation calculator 11. Are multiplied by the optimum feedback gains F ₁ , F ₂ and F ₃ and supplied to the control output. The estimated disturbance _S output by the disturbance estimator 5, the feed forward gain M ₁ is supplied to the control output is multiplied by the arithmetic unit 10. The command value θ _REF is added with the state quantity θ as (−θ), and after passing through the integral element 9, the optimum feedback gain
Is multiplied by F ₄ applied to the control output. In the present embodiment, the control is performed as a continuous system. However, the control system may be a discrete value system by converting the state equation into discrete values and treating them as discrete values. Although the embodiment for the sizing press control has been described above, the effect of this embodiment was confirmed by performing a simulation. The results of examining the followability to the command value by issuing the angle _REF certain command as theta _REF shown in 7-1 Figure-7-3 FIG. FIG. 7-1 is one of the prior arts shown in FIG.
Follow-up performance in case of PI control, FIG. 7-2 shows the case of this embodiment (feedforward control for estimating disturbance is not included), and FIG. 7-3 shows feedforward control for estimating disturbance. Are respectively shown in the case of the present embodiment.
θ _FB indicates the actual value, and the followability to the command value is described in 7-3.
The case of the figure is best, and v _X is also during the pressing (θ _FB =
0.8 to 2.4 [rad]) was almost constant, and the expected control performance was obtained. Next, a comparison was made between the case where the optimal servo problem was made the optimal regulator problem and multivariable optimal control was performed as shown in FIG. FIG. 8-1 shows the case of the conventional method, and FIG. 8-2 shows the case of the present embodiment. In this comparison, T _S = 0. By providing a command value estimator and performing feedback control using the deviation between the measured value and the estimated value of the state quantity, the followability to the command value is clearly improved. In the prior art, since the control is performed such that is 0, interference occurs between the condition that _REF = constant, and the followability is impaired. According to the embodiment described above, the followability to the angle command (constant angular velocity command) is improved as compared with the case where the conventional PI control is used, and the influence of the disturbance is almost eliminated by the feedforward control with respect to the disturbance. In addition to the above-described effects, there is an effect that the followability to the command value is remarkably improved as compared with the case where the conventional optimal servo problem is used as the optimal regulator problem. In the above embodiment, the example of the motor control of the sizing press has been described.
A system that follows a changing target value with a steady-state deviation of 0 (servo system)
It can be applied to all. Instead of describing each embodiment, an apparatus generalizing the present invention will be described below. The state equation of the controlled object is And Here, x is a vector representing an internal state quantity of a control target, u is a vector representing a control output, d is a vector representing a disturbance, and A, B, and E are matrixes of coefficients. It is assumed that the disturbance d is represented by d = f (x) (10). The command value of the servo system is defined as y _REF , the internal state quantity in the equilibrium state corresponding to y _REF is defined as y, and y = Cx (11). y _REF and y are matrices, and C is a matrix indicating coefficients. To make the servo problem a regulator problem, Equation of state (9) becomes Becomes Of the internal state quantities x, those that cannot be detected are determined by forming an observer, and the optimal control gain F (feedback) and M (feedforward) are determined from equation (13). F and M are both matrices. Here, control gain M (feed forward)
Is obtained only for the disturbance d. (The y _REF will be described later.) In the state equation (13), d = 0, far. In the steady state, since y = y _REF ,
seeking x _REF than y _REF, Holds. Here, [x ₀ , u ₀ ] ^T is the internal state quantity and control output in the steady state corresponding to y _REF . The state equation of only x ignoring the disturbance term is as follows: = Ax + Bu (17), and in a steady state, ₀ = Ax ₀ + Bu ₀ (18). By adding u ₀ to the control output u Therefore, = Ax + B (u + u 0) because it is ...... (19) u = -F ( x-x 0), (- 0) = (A-BF) (x- x ₀ ) (20) is obtained. Therefore, u ₀ −F (x−x ₀ ) (21) is output as the feedback control output. When the control system is a discrete value system, the state equation may be converted into a discrete value, and may be treated as a discrete value. The above is summarized as follows. 1) The state equation (9) to be controlled is created, and the servo problem is changed to the regulator problem by using the equation (13). 2) Calculate the optimum control gains F (feedback) and M (feedforward) using equation (13). 3) When the disturbance d is represented by the equation (10), a disturbance estimator is provided, and the estimated disturbance is obtained by the following formula: = f (x). 4) than the command value y _REF of the servo system, the internal state quantity x ₀ in a steady state corresponding to y _REF, provided command value estimator for deriving the control output u ₀ derive the x _0, u _0, control The output, u ₀ −F ′ (x−x ₀ ) −M ·, is calculated and derived. (F 'indicates only the gain for x in F) 5) The output of the steady-state error compensator is u ₀ -F' (x-x ₀ ) -M
・ Add to The general operation of the control device according to the present invention has been described above. [Effects of the Invention] As described above, according to the present invention, based on the state quantity command value, the internal state quantity command value when the state quantity command value changes is estimated, and the estimated internal state quantity command value and the control object The deviation from the internal state quantity is determined, and the control target is controlled based on the deviation and the control output for the control target. Therefore, even if the state quantity command value changes, the followability to the change in the state quantity command value is improved. be able to.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a system diagram showing a basic configuration of the present invention, FIG. 2 is an explanatory diagram of an operation of a sizing press which is an embodiment of the present invention, and FIG. FIG. 4 is a system diagram showing an embodiment in which the present invention is applied to a sizing press. FIG. 5 is a control system diagram in the case of using the conventional PI control. Is a system diagram showing the optimal servo control of the prior art, and FIGS.
FIGS. 8-1 and 8-2 are diagrams showing simulation results of control according to the embodiment of the present invention and the prior art. 1 ... Control target, 5 ... Disturbance estimator, 10 ... Calculator.

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-60-153503 (JP, A) JP-A-61-251902 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) G05B 11/00 G05B 13/00

Claims (1)

(57) [Claims] A deviation between a state quantity command value for a control target having a plurality of internal state quantities and an internal state quantity changing with a change in a state quantity command value and a state quantity actual value indicating an actual state quantity of the control target is input. A steady-state deviation compensating means for generating a control output, a command value estimating means for estimating an internal state quantity command value based on the state quantity command value, and an output of the estimated internal state quantity command value and the control object. A servo, comprising: an internal state quantity deviation calculating means for calculating an internal state quantity deviation from an internal state quantity according to the above; and a control calculating means for controlling a control target based on the control output and the internal state quantity deviation. Control device.