JP3513450B2 - Anti-windup controller - Google Patents

Description

Description: BACKGROUND OF THE INVENTION [0001] 1. Field of the Invention [0002] The present invention relates to an anti-windup controller, and more particularly to an anti-windup controller capable of suppressing a windup phenomenon caused by the presence of a saturation characteristic in a control target. About. 2. Description of the Related Art In a feedback control system, a compensator including an integral operation is used in order to ensure the following performance. However, due to saturation characteristics of an actuator and a drive circuit arranged at a stage subsequent to the compensator, vibration or overshoot called windup is caused. [0003] This phenomenon occurs, for example, when the actuator shifts from a controllable region to a saturation region during control of an actuator or the like, the control target cannot follow the command value. Will continue to maintain the saturation value. Therefore, a deviation occurs between the command value and the fed back detection signal (saturation value), and this deviation is caused by accumulating more and more due to the integration operation of the compensator. [0004] Here, a compensator that has taken such measures is called an anti-windup controller (UWC). There are two types of conventionally known typical anti-windup controllers. 1) Method of stopping integration This is a method of stopping integration operation when the saturation element is saturated, and can be realized by setting the input to the integrator to 0. 2) Method Using Feedback FIG. 2 shows an example of a circuit based on this method of a conventional anti-windup controller. In FIG. 2, in an adder 3, a deviation e 1 between a target value R output from the signal generator 1 and an observation output D detected from a control target (plant) 5.
Ask for. The deviation e1 is input to the adder 7, and the deviation e2 is obtained between the deviation e1 and the feedback signal F1. The deviation e2 is determined by a compensator 9 using PID control or the like.
Is input to In the compensator 9, for example, when the PID control is used, a numerical value 1 obtained by multiplying the deviation e2 by a constant, a numerical value 2 obtained by integrating the deviation e2 with another constant after time integration, and a deviation e2 obtained by a time differentiation After that, another numerical value 3 multiplied by a constant is calculated, and a numerical value 4 obtained by adding all the numerical values 1, 2, and 3 is output to the compensator. The differential operation in the compensator 9 is based on a deviation e.
Since the high frequency component 2 is infinitely amplified, the numerical value 4 is input to the low-pass filter 11. The output of the low-pass filter 11 is input to the control target 5 via the saturation element 13. As described above, an ordinary actuator or a drive circuit of an actuator always has a maximum output value, and cannot output an output exceeding the maximum value. This acts on the system as a saturation factor. From the input of the saturation element 13,
The feedback signal F2 is extracted, and from the output of the saturation element 13,
The feedback signal F3 is extracted. The adder 15 calculates a deviation e3 between the feedback signal F2 and the feedback signal F3. The deviation e3 is multiplied by the reciprocal of the DC gain of the entire compensator preceding the saturation element 13 in the amplifier 17. The output signal of the amplifier 17 is a feedback signal F
It is negatively fed back to the adder 7 as 1. In FIG. 2, the range shown by the dotted line is the anti-windup controller 10.
It is. At this time, since the input of the integration circuit of the compensator 9 converges to 0, it has a function of stopping the integration operation after the convergence. However, the conventional 1)
The methods described in (2) and (3) have the following disadvantages.
Although the method 1) is considered to be the most effective in principle, a discontinuous operation is performed because the deviation e, which is an integrator input, is suddenly assumed to be 0. Also, when the integration is started again, the input of the integrator is suddenly changed from 0 to the deviation e, so that a discontinuous operation is also performed. The result of the integration function is that both of them change continuously, but the finally obtained output of the compensator includes this integration result. It is known that it is desirable for the control input, that is, the compensator output, to change as continuously and smoothly as possible so as not to induce vibration in the controlled object. Therefore, the method 1) has disadvantages. . On the other hand, the method 2) has the characteristic of compensating for the disadvantage of 1). When the saturation element saturates, the integral input converges asymptotically to 0, and when it is released from saturation, it also asymptotically. It has the function of converging on the deviation e. However, in this method, the integrator continues the integration operation until the integration input converges to 0, so that the integration has an excessively large output value when the integration is stopped. As mentioned above, since the integration result is included in the compensator output,
The output of the compensator also becomes an excessive value. From this, the windup phenomenon is improved but not eliminated. The present invention has been made in view of such conventional problems, and has as its object to provide an anti-windup controller capable of suppressing a windup phenomenon caused by the presence of a saturation characteristic in a control target. . According to the present invention, there is provided an anti-windup controller for performing feedback control of a control target having a saturation characteristic when a control target exceeds a predetermined control range. , A first deviation output means for feeding back the value of the state quantity detected by the state quantity detection means, comparing the value with a target value, and outputting a first deviation, and the first deviation output Means for comparing the first deviation outputted by the means with the feedback signal and outputting a second deviation; and a second deviation output means outputted from the second deviation output means.
A compensating means for performing an operation including integration with respect to the deviation and outputting a compensation signal; a filter for removing a high-frequency component from the compensation signal output from the compensating means; After being restricted based on the characteristic, a saturated element input to the controlled object, a third deviation output means for comparing an output signal of the saturated element with an output signal of the compensation means and outputting a third deviation, Amplifying means for multiplying the output signal of the third deviation output means by a constant multiple and / or a predetermined transmission element and outputting the feedback signal. The state quantity detecting means detects the state quantity of the control object. The state quantity is, for example, temperature or pressure. First
The deviation output means feeds back the value of the state quantity detected by the state quantity detection means and compares it with a target value to output a first deviation. The second deviation output means compares the first deviation outputted by the first deviation output means with a feedback signal and outputs a second deviation. The compensating means performs an operation including integration on the second deviation output from the second deviation output means, and outputs a compensation signal. The compensation means is, for example, PID control. The filter removes high frequency components from the compensation signal output from the compensation means. In the saturation element, the output signal of the filter is input to the control target after being restricted based on a predetermined control range and saturation characteristics. The third deviation output means compares the output signal of the saturation element with the output signal of the compensation means and outputs a third deviation. The reason why the output signal of the filter is compared with the output signal of the compensating means is that, when the output signal of the filter is compared with the conventional one, a phase delay occurs due to the transfer characteristic of the filter. This is because the convergence of the deviation is expected to be delayed accordingly. The amplification means multiplies the output signal of the third deviation output means by a constant multiple and / or a predetermined transmission element, and outputs a feedback signal. By passing through the transmission element, it is possible to further speed up the convergence of the deviation. As described above, the responsiveness of the control system is improved, and the control system has a function of improving safety. Also, compared to the conventional anti-windup controller, the input / output signal of the controller changes stably and smoothly, and the transition from the saturated state to the non-saturated state occurs because the input of the saturation element does not greatly exceed the saturation value. Is performed more smoothly in a short time. Embodiments of the present invention will be described below. The embodiment of the present invention shows an example in which the present invention is applied to temperature control in a refrigeration system. The refrigeration system is, for example, a case in which the temperature in a vehicle compartment is controlled using a variable capacity gas compressor. FIG. 1 shows a configuration diagram of an embodiment of the present invention. Note that the same components as those in FIG. 2 are denoted by the same reference numerals and description thereof is omitted. In FIG. 1, the control target 5 is a quaternary system or the like.
It is approximated by a higher-order transfer function. The feedback signal F4 is extracted from the input of the low-pass filter 11 and input to the adder 15. In the adder 15, the feedback signal F4 and the feedback signal F4
The deviation e4 of 3 is calculated. The deviation e4 is calculated by the amplifier 27 using the D
The adder 7 is multiplied by the reciprocal of the C gain and generates a feedback signal F1.
Is fed back to Alternatively, a transmission element may be added in addition to the DC gain to cancel a phase delay caused by passing through the compensator 9 or the like. The deviation e4 becomes 0 in a range where the input value of the saturation element 13 does not exceed the saturation value. When the input value exceeds the saturation value, the feedback signal F3, which is the output of the saturation element 13, is fixed at a constant saturation value, but the other feedback signal F4 is not fixed. The difference e4 between the two is negatively fed back to the input of the adder 7 via a constant multiple or via a transfer function, so that the input to the compensator 9 converges to zero. At this time, since the feedback signal F4 is extracted from the input of the low-pass filter 11, the signal can be fed back without any phase delay due to the low-pass filter 11. In FIG. 1, the range shown by the dotted line is the anti-windup controller 20. By adopting the above configuration, the anti-windup controller 20 has a function of suppressing a windup phenomenon caused by the presence of the saturation element 13, so that the physical quantity to be controlled changes excessively, Is to be suppressed. As a result, the responsiveness of the system is improved, and it also has the function of improving safety. Further, as compared with the conventional anti-windup controller, the input / output signal of the controller changes more stably and smoothly, and the input of the saturation element does not greatly exceed the saturation value. The transition to is performed more smoothly in a short time. In a refrigeration system having an external capacity control type compressor in a car air conditioner, the heat load fluctuates greatly. At a high heat load, the capacity is saturated at the maximum capacity, and at a low heat load, the capacity is saturated at the minimum capacity. Since such saturation occurs very frequently, the present invention is particularly useful for maintaining system stability. Further, the present invention can be applied to mechatronic devices such as robots, machine tools, steppers, hard disk drives, DVDs, etc., and all devices in which feedback control such as temperature control and pressure control is used in scientific plants and refrigeration systems. Things. The operation algorithm in the anti-windup controller 20 is realized as a program running on a computer or by an electronic circuit. In the former case, the operation algorithm is stored in a recording device such as a memory circuit, a disk, a CD, and a DVD. As described above, according to the present invention, the third deviation output means is configured to compare the output signal of the saturation element with the output signal of the compensation means. Has the effect of improving safety.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an embodiment of the present invention. FIG. 2 is a circuit example of a conventional anti-windup controller. [Description of symbols] 1 Signal generators 3, 7, 15 Adder 5 Target 9 Compensator 20 Anti-windup controller 11 Low-pass filter 13 Saturation element 27 Amplifier R Target value D Observation output F1, F3, F4 Feedback signal e1, e2, e4 Deviation

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-2-58102 (JP, A) JP-A-62-236002 (JP, A) JP-A-6-203498 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) G05B 11/36 B60H 1/32

Claims (1)

(57) [Claim 1] An anti-windup controller for performing feedback control of a control target having a saturation characteristic when exceeding a predetermined control range, wherein a state amount of the control target is detected. Quantity detection means, first state output means for feeding back the value of the state quantity detected by the state quantity detection means, comparing the value with a target value, and outputting a first deviation, and output by the first deviation output means The second deviation output means for comparing the first deviation with the feedback signal and outputting a second deviation, and performing an operation including integration on the second deviation output from the second deviation output means, thereby compensating the second deviation. A compensating means for outputting a signal, a filter for removing high-frequency components from the compensating signal output from the compensating means, and the control target Entered in A third deviation output means for comparing an output signal of the sum element and the output signal of the saturation element with an output signal of the compensating means and outputting a third deviation; Or an amplifying means for applying a predetermined transmission element and outputting the feedback signal.