JPS6217242B2 - - Google Patents

Description

[Detailed description of the invention]

<Industrial Application Field> The present invention relates to a process adaptive control device. <Conventional technology and its problems> In conventional equipment that performs process control so that the controlled variable matches the target value r, it is difficult to set the adjustment value so that the optimum control result can be obtained around the rated load of the process. However, as the load decreases, the change in the control amount increases, which poses a problem in process control. Adaptive control can address the above-mentioned problems by detecting process characteristics and adaptively and automatically adjusting controller adjustment values, but traditional adaptive control relies on computers to detect process characteristics. Alternatively, by detecting other process variables that vary with the load, generating an adaptive signal, and providing a mechanism to force each controller adjustment value to change in correlation to the magnitude of the adaptive signal. The drawback is that the device is complicated and expensive. This invention solves the above-mentioned conventional problems, and when the load changes, each adjustment value of the controller is automatically corrected in accordance with the changed load, without using complicated mechanisms or computers, etc. The purpose of the present invention is to provide an adaptive control device for a process that can always obtain optimal adjustment results. <Means for solving the problems> In order to achieve the above object, the present invention provides a load v
In order to match the controlled variable of the process to the target value, the controller has proportional, differential, and integral elements arranged in this order, and the characteristics (gain R, time constant T, dead time L) at the rated load of the process are The controller's proportional, differential, and integral adjustment values are set so that the evaluation function takes an extreme value based on A square root element is inserted after the integral element at the final stage of the controller. Then, by multiplying and dividing the output signal √ by each of the proportional, differential, and integral elements within the controller, each adjustment value of this controller with respect to a change in load v can be calculated as follows: The characteristics (gain T〓, time constant R〓, dead time L〓) are also characterized by being configured so that the evaluation function is automatically converted to an extreme value. <Operation and Principle of the Invention> The present invention has been made in view of the following points in an apparatus that performs adaptive control, that is, detects process characteristics and automatically adjusts adjustment values of a controller. () The size of the regulating valve is generally chosen to be twice the process rated load v. Therefore, if the control circuit is designed so that the manipulated variable m is expressed by the following formula: m=v1+ _2LS /1+ _LS ...(1), the control area will be minimized, or in other words, the evaluation function will be at its extreme value, and the controller's The optimum adjustment value can be set as a function of the characteristics R, T, and L at the process rated load. Here, R is the gain, T is the time constant, L is the dead time, and s is the differential symbol d/dt
It is. () It was confirmed through experiments that the process characteristics R, T, and L change in inverse proportion to the square root of the load v applied to the process. That is, it was found that if the characteristics R〓, T〓, L〓 when the load changes, the following relationship exists with the characteristics R, T, L at the rated load. This invention utilizes the relationship between (1) and (2), and when the controller's proportional, differential, and integral adjustment values are set so that the process evaluation function takes an extreme value, the controller Since the output of the integral element at the final stage of is a signal corresponding to the load v, lead this to the square root element,
If the output signal √ is multiplied by each proportional, differential, and integral element in the controller, each adjustment value of the controller is determined by the relationship of equation (2), which is the characteristic R〓, T〓, L 〓 is also automatically converted so that the evaluation function becomes an extreme value. In other words, the manipulated variable after the load change is It is converted into . <Example> Next, an example of the present invention will be described based on the drawings. The manipulated variable m is v1+2L _S /1 for a first-order delayed process (gain R, time T) including dead time L.
If the control circuit is designed to provide + _LS , the result will be as shown in Figure 1. The part indicated by the chain line 11, that is, the dead time L
The case where the rated load v is applied to the primary delay process including Note that since the dead time L is transiently equivalent to a first-order lag where the time constant is L, the dead time e ^-LS is expressed as a first-order lag. chain line 12
The controller shown is proportional P, differential D, integral I
A control circuit is constructed with series operation, and its optimal adjustment values are a proportional gain of 2T/RT, a differential time L, an integral time T, and 2L. FIG. 1 is shown as the following formula. -(m1/1+L _S -v)R/1+T _S・2T/RL(1+L _S )(1+1/T _S )(1+1/2L _S
)=m...(4) If we rearrange this equation (4) and find m, we get the following equation. m=v1+2L _S /1+L _S ...(5) This equation (5) is the signal 1v1/1+L of the controller 12.
_S and signal 2v1+2L _S /1+L _S ,
Further, signal 1 is a first-order delayed form of v, and is a signal that becomes v after a short period of time has passed. Therefore, by square rooting signal 1, √ in equation (2) can be derived. Next, in an embodiment of the present invention, the configuration and operation of the control circuit will be described when the load v applied to the process is reduced from the rated load and the process characteristics become R〓, T〓, L〓. Since the adjustment values of the controller 22 are the set values when the load is constant, it is necessary to change the values to R〓, T〓, L〓 and make the manipulated variable m according to equation (3). Therefore, as shown in FIG. Multiply each proportional, differential, and integral element. Substitute equation (2) into equation (6) to find m. Below is the result of the calculation

[Formula] becomes. In this way, by inserting √, the characteristics R, T, and L are converted into R〓, T〓, L〓, and the manipulated variable m of equation (7) that yields the optimal adjustment result can be found. . Next, in the case of integral property (integral time constant T) including dead time L of the process, a control circuit shown in FIG. 3 may be designed. A signal 6 of √ obtained by transmitting the signal 5 of the controller 32 through the square root element H
is introduced into the multiplication element V and the division element W. Then, the control system shown in FIG. 3 is expressed by the following relationship. Substitute equation (2) into equation (8) to find m. Below is the result of the calculation

[Formula] becomes. Since this equation (9) matches equation (7), the optimum control result can be obtained in the same way as when the load v is reduced from the rated load. <Effects of the Invention> As explained above, according to the present invention, by utilizing the fact that equation (2) holds true in most processes, a signal corresponding to the load v extracted from the controller is converted into a square root element. By introducing the signal √ into the controller and multiplying and dividing the signal √ by each of the proportional, differential, and integral elements in the controller, the controller adjustment value can be autonomously evaluated against the process characteristics after a change in the load v. The function is modified to take an extreme value, and as a result, optimal control can always be achieved even when the load v varies. Moreover, unlike conventional adaptive control, there is no need for complex mechanisms or expensive computers; instead, it is only necessary to add elements with square root and multiplication or division functions within the controller, resulting in a simple and inexpensive adaptive control device. be able to.

[Brief explanation of the drawing]

FIG. 1 is a block diagram for explaining a control system according to the present invention. FIG. 2 is a block diagram showing the configuration of the above embodiment. The third is a block diagram showing the configuration of another embodiment of the invention. 11, 21... Primary delayed process including dead time, 12, 22, 32... Controller, 31
... Integral process including dead time, r... Target value, m... Manipulated amount, v... Load, R... Gain,
L...Dead time, T...Time constant, V...Multiplication element, W...Division element.

Claims (1)

[Claims] 1 In order to match the controlled variable of the process on which the load v acts, with the target value, in a controller in which proportional, differential, and integral elements are arranged in this order, the characteristics of the process at the rated load (gain R, time constant T, With the proportional, differential, and integral adjustment values of the controller set so that the evaluation function takes an extreme value based on the dead time L),
Taking advantage of the fact that the output of the integral element at the final stage of the above controller becomes a signal corresponding to the load v, a square root element is inserted after the integral element at the final stage of the controller, and the output signal √ is applied to the controller. By multiplying and dividing the proportional, differential, and integral elements of An adaptive control device for a process, characterized in that the evaluation function is automatically converted to an extreme value even for dead time L).