JP2707075B2 - Plant simulation equipment - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a plant simulation apparatus, and is particularly suitable for creating a model necessary for designing a multivariable control system and analyzing the model. It relates to a plant simulation device. [Prior art] Conventional plant simulation equipment is described in "Tanuma, Morooka, and Takano," Interactive CAD System for Control Systems, "Hitachi Review, No. 6.
As discussed in "Vol. 5, No. 3, March 1983", it has been proposed as an analysis system that can be represented by linear ordinary differential equations such as first-order lag elements, second-order lag elements, and dead time elements. By the way, it is very difficult to express the rolling phenomenon and the like using only the elements which can be analyzed as employed in the above-mentioned plant simulation apparatus. The model for designing the system was created by approximating it with linear elements from what was expressed in a computer language (for example, FORTRAN) procedure. The control system was designed using the created model.However, the system designed using the model created by the linear approximation in this way includes an error in the model, so that the operation state of the actual system Therefore, in a system designed using the model, the constants of the control device were adjusted after installation at the site to obtain a desired system response, however, in the case of a system such as a steel rolling plant, In order to adjust the constants, the trial rolling requires a large amount of material, so that it is not possible to perform the trial rolling in which the constants are changed to a large extent. Range that does not significantly affect the rolling results at the time
That is, since adjustment is performed to change the constant of the control device within a range in which the material to be rolled can be shipped as a product, a large amount of time is required for adjustment. In addition, as described in Japanese Patent Application Laid-Open No. 61-3202, a model in which a model of a plant to be controlled is obtained from an input / output data to a control system by an autoregressive method has been proposed. As described in Japanese Patent Application Laid-Open No. 61-267103, a method of collecting information on a plant in order to create a model of the plant, or using a knowledge base or the like to create a model Has been proposed. [Problems to be Solved by the Invention] The above-mentioned conventional technology accurately reflects the input / output relationship of the controlled object in the model, or configures the model assuming that the controlled object is a nonlinear element, and corrects the parameters of the model. However, even if a control system is designed using a model, and a system including this control system is installed and adjusted, the system must be adjusted again after the system is installed. ing. An object of the present invention is to accurately describe a model of a control target,
An object of the present invention is to provide a plant simulation apparatus capable of eliminating readjustment of a system after installing a system for controlling a plant. [Means for Solving the Problems] In order to achieve the above object, the present invention provides an operation amount based on an input to a control device, an output of the control device in response to the input, and an output of a plant in response to the operation amount. And a block element for outputting information corresponding to an operation amount by an output of the control device in response to the input stored in the storage device, representing a function of the control device as a parameter. And expressing the operation of the plant as a parameter excluding a time function element, in response to the manipulated variable stored in the storage means and the output of the block element means, and outputting information on an operation similar to the operation of the plant. And outputting an error signal indicating an error between the output of the model means and the output of the plant stored in the storage means. Comparing means for comparing the output of the block element means with the operation amount stored in the storage means to output an error signal indicating an error between the two, and each error signal output from the comparing means being minimized As described above, a plant simulation apparatus includes the parameter changing means for changing the parameters of the model means and the parameters of the block element means. [Operation] According to the means described above, the input information of the control device is given to the block element means having the function of the control device for controlling the plant, and the manipulated variable by the output of the control device is output to the model means simulating the plant. The output of the block element means is given, the output of the block element means is compared with the manipulated variable based on the actual output of the control device, and the output of the model means is compared with the actual output of the plant. Since the parameters of the model means and the block element means are respectively changed, the control device can be designed in the same manner as that for an actual plant, and a system for controlling the plant After the installation, the system only needs to be adjusted once in consideration of the input / output relationship of the plant, and the maximum adjustment of the system is not required. Embodiment An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of the present invention. In FIG. 1, 1 is a plant, 2 is a control device, 3 is a data collection device, 4 is an analog / digital (A / D) converter, and 10 is a plant simulator. The plant simulator 10 includes a memory 11 as storage means, and a model means.
12, a comparator 13, a parameter changing means 14, and a block element means 15. The operation of the embodiment having the above configuration will be described. A plant 1 such as a steel rolling mill is controlled by a control device 2.
Control is performed so that a desired output is obtained. That is, when the input S ₁ is input to the control device _2, an output of the operation amount S ₂ is obtained from the control device 2. When the operation amount S ₂ are input to the plant 1, it affects the output S ₄ states S ₃ and plant 1 of the plant 1. These input S ₁ , manipulated variable S ₂ , state S ₃ , and output S ₄ are input to the data collection device 3. The data collection device 3 may be a simple level converter, or may be of a type that is recorded on a magnetic tape like a data recorder and can be used at a later date. When the signal from the data collection device 3 is an analog signal, the output from the data collection device 3 is converted into a digital signal by the A / D converter 4 and stored in the memory of the plant simulator 10.
Stored in 11. If the output of the data collection device 3 is a digital signal, it is stored in the memory 11 immediately after. Among the data stored in the memory 11, data corresponding to the operation amount S ₂ of the plant 1 is input into the model unit 12. The model means 12 is obtained by simulating the plant 1, and the output of the model means 12 is inputted to a comparator 13. Wherein the comparator 13, the information corresponding to the output S ₄ of the plant 1, which is stored in the memory 11 is input. The comparator 13 obtains an error by comparing the information corresponding to the output S ₄ and the output from the model unit 12, the error signal is input to the parameter changing means 14. The parameter changing means 14 is for a person familiar with the plant 1 to change the parameters of the model means 12 based on the error signal. Here, a person familiar with the plant 1 (referred to as an expert) uses the parameter changing means 14 to change the parameters of the model means 12 so that the error signal is equal to or less than the allowable error. When the error signal is less than the allowable error, the signal stored in the memory 11 corresponding to the input S ₁ of the control device 2 is inputted to the block element means 15. The output of the block element means 15 and the operation amount of the memory 11 are input to the comparator 13, and the error signal, which is the output signal of the comparator 13, is input to the parameter changing means 14, and the error signal is reduced to zero by the parameter changing means 14. The parameters of the block element means 15 are changed so that As a result, the plant 1 is equivalent to the model means 12 and the control device 2 is equivalent to the block element means 15. Here, the control device 2 can be expressed by a combination of linear elements such as a PID control device or the like and simple nonlinear elements such as limiters and switches, and corresponds to the block element means of the simulator 10 and has the same parameter value. Normal. When the adjustment of the parameters of the block element means 15 and the model means 12 is completed, the input signal of the memory 11 is inputted to the block element means 15. The output from the block element means 15 (in the actual plant, the signal corresponding to the operation amount S ₂₎
Is input to the model means 12. The model means 12 is
Receiving a signal corresponding to the operation amount S _2, and outputs a signal corresponding to the output S ₄ to comparator 13. Further, the comparator 13 compares the output from the model means 12 with the plant output stored in the memory 11, and
Is compared with a signal corresponding to the operation amount stored in the memory 11, and an error signal of the comparison result is output to the parameter changing means 14. The parameters of the block element means 15 and the model means 12 are changed using the parameter changing means 14 so that each error signal is equal to or less than the allowable error. As a result, response characteristics equivalent to those of the control device 2 and the plant 1 are obtained by the plant simulator 10. The simulator 10 that has completed the parameter adjustment operates simultaneously with the control device 2 and the plant 1, so that the external input S ₀ according to various operating states including signals that cannot be actually input in the actual plant is stored in the memory 11. Various simulations can be performed by inputting signals instead of input signals. Therefore, according to the present embodiment, there is no need to design a system and perform readjustment after installation of the system. Next, regarding the specific processing of the plant simulator 10,
This will be described with reference to FIG. As an example of a plant, in a steel rolling mill system,
In a speed control system for controlling the speed of a roll or a reduction system for controlling the gap of a roll, the system is continuous with respect to time, and as the control device 2, a linear control element of a first-order lag element or a second-order lag element, a limiter, It can be represented by a non-linear element such as a switch. These elements represent differential equations. For example, in the case of a first-order lag element as shown in FIG. Can be represented by the differential equation At the same time, the second-order lag element can be described as a differential method. In general, the control device 2 is a PID control device that can be described by the above-described first-order delay element, second-order delay element, and the like. As a result, it can be described by a block diagram. On the other hand, if the plant 1 to be controlled is a rolling mill, it depends on the rolling equation of HILL as shown in the following equation (2). In this case, since equation (2) is represented nonlinearly, it is difficult to analytically find a solution by combining the linear elements. On the other hand, in a multi-stage rolling system, etc., the speed control system and the reduction system are controlled by an analog amount, but in a multi-high rolling mill, the speed command and cage command for each stand are issued from a digital computer or the like, so-called discrete value system. It is often done with a control system configuration. FIG. 3 shows a process for obtaining the time response of a system including a linear element, a nonlinear element, and a discrete element. The processing shown in FIG. 3 is described in “Miwa Masatake et al.,“ Iwanami Course Information Science-18 Numerical Calculations ”, Iwanami Shoten, January 1982, pp. 144-145.
This is an application of the Runge-Kutta formula described on page ". In other words, when the first-order lag elements and the like are represented by a block diagram, these elements are mechanically expressed by Equation (1). Is expressed by a differential equation (simultaneous first-order ordinary differential equations), and a continuous system for obtaining f (x ₁ , x ₂ ...) At a certain time is calculated (step 100). Calculate a model system (continuous) that calculates a part that can be expressed by a continuous system, for example, the HILL formula as shown in equation (2) (step 11).
0, steps 100, 110). Next, Runge-Kutta calculation (RK calculation) for finding the equivalent of the cracking time is performed (step 120). It is determined in step 130 whether or not the calculation of RK has been performed four times. If less than four, the process proceeds to step 100, and if the calculation of RK is completed, the process proceeds to step 140. When the time corresponding to the sampling period of the discrete value system is reached, the process proceeds to step 150, but otherwise, step 100 is executed (step 140). A discrete value system is calculated (step 150), and then a discrete part of the model system is calculated (step 160). Then, it is determined whether or not to end (Step 170). Here, the calculation step 150 of the discrete value system will be described using an example. For example, x (k + 1) = ΦX (k) + ΨU (k) (3) y (k) = CX (k) + DU (k) (4) ,
The state X (k) of the k-th step and the input U (k) are stored in the memory, and the output is obtained by multiplying X (k), U (k) by the output matrix [CD] from the right. The state of the step is obtained by using the calculation on the right side of Expression (3), and the obtained X (k + 1) is substituted for X (k). By such an operation, the state of the next step is sequentially obtained from the current state and the input. On the other hand, the parameter changing means 14 can set items other than h of the HVS output, which is the input of the HILL rolling equation of the formula (4), as parameters. For example, when the expression (4) is described in a FORTRAN program, it becomes an initial value. Therefore, this can be realized by adding a program for setting the initial value. With the above-described configuration, the steel plant can obtain various coefficients and control it with high precision in accordance with the rolling phenomenon in the high-speed region. Then the accuracy will drop. This is because iron generated when operated in a high-speed region has high accuracy and can be shipped as a product, and data collection is easy. Therefore, it is possible to adjust parameters while collecting data. However, in the low speed range of about 1/100, parameters such as friction coefficient and oil film thickness are so different that iron products with the accuracy that can be shipped as products cannot be produced. Data was being collected. The data is stored in the memory of the simulator 10, the difference between the actual model and the model is obtained, and the parameter is slightly changed to observe the change in response. As a result, an error signal output from the comparator 13 fluctuates. The parameters are corrected so that the shake is minimized and the value of the error signal is minimized. The system is adjusted by this repetition. When the model 12 faithfully simulates the actual system, the control device is designed and the block elements are constructed so as to obtain the desired response, and the response is obtained and corrected. As a result, the correction result of the control device can be simulated without using a local rolling mill, and the development period can be greatly reduced. FIG. 4 is a block diagram showing a modification of the present invention.
In the embodiment of the present invention, an optimum countermeasure for changing a parameter is obtained by using the knowledge of a plant expert. That is, the difference between the embodiment shown in FIG. 4 and the embodiment shown in FIG. 1 is that the error signal of the comparator 13 is inputted to the inference means 16, and the inference means 16 determines the parameter to be changed using the knowledge base 17. That is, the parameter to be changed is instructed to the parameter changing means 14. An example of the knowledge base 17 is shown in FIG. For example, in the case of iron sheet rolling, when the rolling speed is reduced, the friction coefficient is reduced as a parameter, and as a phenomenon, the generated sheet thickness is increased. Using this knowledge, the present embodiment executes the inference program shown in FIG. That is, the condition part of the rule is extracted (Step 200). Next, a condition part matching the departure state is found (step 210). Further, the matched condition part is removed and replaced with the conclusion part (step 220). It is determined whether the state matches the target state (step 230). With this operation, when the rolling speed is reduced, when the phenomenon that the generated sheet thickness becomes thicker occurs, when the phenomenon that the rolling speed decreases occurs, the state of the phenomenon is removed from the state that the speed is reduced, and the generated sheet thickness is reduced. It replaces the phenomenon of thickening. The phenomenon that the error signal generation plate thickness becomes large occurs, which coincides with the inferred phenomenon. As a result, the parameters are changed by obtaining the conclusion of the operation unit that reduces the friction coefficient. In this way, by incorporating the knowledge of the qualitative expression expert, the model can be refined. Note that, instead of the inference means 16 and the knowledge base 17 shown in FIG. 4, an identification mechanism 18 as shown in FIG. 7 may be provided. The processing algorithm in this case is described in “System Identification”, et al.
Page to page 259 ". In this case, unlike the inference method, it is not ambiguous information, and the parameter can be quantitatively changed. As described above, according to the present invention, input information of the control device is given to the block element means having the function of the control device for controlling the plant, and the operation amount by the output of the control device is given to the model means simulating the plant. And the output of the block element means, and the output of the block element means is compared with the manipulated variable based on the actual output of the control device, and the output of the model means is compared with the actual output of the plant. Since the parameters of the model means and the block element means were changed respectively, the control device was designed for an actual plant. That's and can be equally carried out, after the setting of the system for controlling the plant, it is possible to eliminate the re-adjustment of the system.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a block diagram showing a display example of a first-order lag system, and FIG. FIG. 4 is a block diagram showing another embodiment of the present invention, FIG. 5 is an explanatory diagram showing an example of a knowledge base, FIG. 6 is a flowchart showing an inference algorithm, and FIG. FIG. 13 is a configuration diagram showing another embodiment of the present invention. 11 ... memory, 12 ... model means, 14 ... parameter changing means, 15 ... block element means, 16 ... inference means, 17
…… Knowledge base, 18 …… Identification mechanism.

   ────────────────────────────────────────────────── ─── Continuation of front page    (72) Inventor Harumi Maruyama               4026 Kuji-cho, Hitachi City Hitachi, Ltd.               Inside the Hitachi Research Laboratory (72) Inventor Goo Shimizu               5-2-1 Omikacho, Hitachi City Stock Association               Omika Plant of Hitachi, Ltd. (72) Inventor Masahiro Tobiyo               5-2-1 Omikacho, Hitachi City Stock Association               Omika Plant of Hitachi, Ltd.                (56) References JP-A-61-3202 (JP, A)                 JP-A-61-267103 (JP, A)                 JP-A-61-65307 (JP, A)                 JP-A-61-199023 (JP, A)

Claims (1)

(57) [Claims] Storage means for inputting and inputting an input to the control device, an operation amount based on an output of the control device in response to the input, and an output of the plant in response to the operation amount, and storing the function of the control device as a parameter A block element means for outputting information corresponding to an operation amount by an output of the control device in response to an input stored in the means, and an operation of the plant represented by a parameter excluding a time function element and stored in the storage means Model means for outputting information on an operation similar to the operation of the plant in response to the output of the manipulated variable and the output of the block element means, and an output of the model means and an output of the plant stored in the storage means. An error signal indicating the difference between the two is output, and the output of the block element means is compared with the operation amount stored in the storage means. Comparing means for outputting an error signal indicating an error between the two, and changing the parameters of the model means and the parameters of the block element means so that each error signal output from the comparing means is minimized. And a means for plant simulation.