JPH02240703A - Programmable controller - Google Patents

Abstract

PURPOSE:To attain the fuzzy control with an optional input number and an optional rule number by preparing an instruction for the fuzzy control that can be described together with a ladder diagram and setting optionally the input number and the rule number of the fuzzy control which is carried out by the instruction for the fuzzy control. CONSTITUTION:A CPU 1 is used together with a system program memory 2, a user program memory 3, an I/O 4, and a system bus 5. A fuzzy control instruction is prepared to carry out the fuzzy inference having a form suited to be described together with a ladder diagram type program. Thus the output of the fuzzy control instruction can be combined with another instruction having another input and control rule. Thus it is possible to describe the fuzzy control having an optional input number and an optional rule number with combination of the simple fuzzy control instructions having a small number of inputs and rules respectively.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a programmable controller used for controlling industrial systems such as steel, paper manufacturing plants, public systems, and automobile manufacturing, and particularly relates to This invention relates to a programmable controller in which fuzzy control and control instructions are added to the sequence instructions that are the core of the control function.

(Prior Art) In recent years, fuzzy control has been used for control in fields where it is difficult to apply PID control or modern control theory.

The first application was to control cement and kilns, but the field of application has expanded not only to process control and operation control, but also to pattern recognition and artificial intelligence.

Fuzzy control, like an expert system, uses the knowledge of a human operator as If'-then type control rules. However, the feature is that a fuzzy set can be written in the condition part and the conclusion part of the 1['-then rule.

However, the only hardware that performs the above control is a dedicated computer using a fuzzy computer or fuzzy LSI, or a dedicated controller specially manufactured for the controlled object, and there has been no example of application to a programmable controller.

(Problems to be Solved by the Invention) As described above, conventionally there has been no programmable controller capable of fuzzy control. Furthermore, there was no established method for expressing fuzzy sets, which are essential for fuzzy control, on a programmable controller.

The present invention allows fuzzy control to be described in a way that is compatible with the ladder diagram type (Figure 9) program used in ordinary programmable controllers, and allows the number of human forces and rules for control to be set arbitrarily. The purpose of this invention is to provide a programmable controller with fuzzy control instructions.

[Structure of the invention]

(Means and Effects for Solving the Problems) The present invention provides a programmable controller programmed using a ladder diagram, in which a fuzzy control instruction for executing fuzzy inference in a form suitable for being written in combination with a ladder diagram type program is provided. Fuzzy control with any number of inputs and rules is possible by preparing the fuzzy control command and configuring it so that the output of the fuzzy control command can be used in combination with other human power and fuzzy control commands with control rules. It is a programmable controller.

(Example) FIG. 1 shows the hardware configuration of an embodiment of the present invention.

In FIG. 1, 1 is a CPU, 2 is a system program memory, 3 is a user program memory, 4 is Ilo,
5 is a system bus. FIG. 2 shows an embodiment of the fuzzy instruction of the present invention.

In FIG. 2, 20 is the instruction body, 21 is the condition input, and 22
is intermediate output register clear input, 23 is execution output, 24
25 is an input register display, 25 is an instruction name, 2B is a rule storage register display, and 27 is an output register display. An example of how this command is used is shown in FIG.

First, the fuzzy inference method will be explained. Now, as rule 1, H(Xl-A) and (X2-B)
Consider the rule then (y-C:). Here, Xl and X2 are independent inputs, y is an output, A and B represent a fuzzy set of the conditional part, and C represents a fuzzy set of the conclusion part. The meaning of rule 1 is that if X is A and X2 is B, set the output y to C. FIG. 4 illustrates this situation. In Figure 4, A,
The curves shown by B and C (triangles here) show the membership functions of the fuzzy set.

For example, the value αx1 corresponding to the input value X 1 indicates the extent to which xl belongs to the fuzzy set A, that is, the membership value. The membership value has a value from O to 1. Similarly, for the input X2, β8□ indicating the B-likeness is obtained. To obtain the output, cut out the output fuzzy set at the minimum value (MINei calculation) of αx1 and βX2 (in this case, the shaded area in the diagram showing the output in Figure 4), and correspond to the position of the center of gravity of this part. The value of y becomes the output. If there are multiple control rules, the fuzzy sets of the conclusion parts of each are logically summed, and the value corresponding to the center of gravity is output. This situation is shown in FIG.

In the programmable controller of the present invention, the membership functions of the conditional part and the conclusion part are stored in registers as numerical values. In other words, if the input resolution is, for example, 4 bits,
It is assumed that the input is divided into 16 stages and the membership values corresponding to the input values of each stage are written in 16 registers. This situation is shown in FIG. If other fuzzy sets are expressed in the same way, the instruction shown in FIG. 2 will require the registers shown in FIG. In other words, the input register is 2 words, and the description of the fuzzy set in the condition part is 16 words + 16 words.
32 words, and the conclusion section also has an output resolution of 4
16 words in bits, plus register 1 for output
word. The actual calculation contents of this instruction are shown in the flowchart of FIG.

The present invention is characterized in that it has a register for intermediate outputs in addition to the fuzzy set description registers in the condition part and conclusion part described above. This register may be located at a fixed address, so there is no need to display the address on the instruction. This intermediate output register is a register that stores the values of the shaded part in the fuzzy set of the conclusion part in FIG. If you want to save the graph shape of the fuzzy set in the conclusion section in Figure 4 and obtain outputs based on multiple control rules, you can use this intermediate output when executing fuzzy instructions with other control rules. By repeatedly using the contents of the register while calculating the logical sum, results according to a plurality of control rules as shown in FIG. 5 can be obtained.

If an output based on an independent control rule is required, a new output is required by clearing the contents of the intermediate output register. To this end, the fuzzy instruction of the present invention is provided with an intermediate output register clear input.

By turning on this input, the intermediate output register is cleared. In this way, it is possible to select whether to use an independent control rule or to combine it with other control rules each time an instruction is executed.

In the above explanation, in order to simplify the explanation, a fuzzy control instruction with two human inputs, one rule, and one output was used. However, fuzzy control instructions with arbitrary numbers of inputs, rules, and outputs can also be used. It goes without saying that they can be used in combination using the same method.

[Effects of the Invention] As explained above, according to the programmable controller having fuzzy instructions of the present invention, by using a combination of simple fuzzy control instructions with a small number of inputs and rules, any number of inputs can be controlled. , fuzzy control of the number of rules can be described. Furthermore, it has a form that can be easily used in a programmable controller that uses a ladder diagram, and the description of fuzzy sets is very easy, so it has the effect of being very easy to apply to actual use.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the hardware configuration of the programmable controller according to the present invention, FIG. 2 is a diagram showing an example of fuzzy control instructions used in the programmable controller according to the present invention, and FIG. 3 is a block diagram showing the hardware configuration of the programmable controller according to the present invention. A diagram showing a program example of fuzzy control instructions, FIG. 4 is a diagram to explain fuzzy inference, FIG. 5 is a diagram to explain a method for obtaining fuzzy control output using multiple control rules, and FIG. 6 7 is an explanatory diagram of the register used for the fuzzy control instruction according to the present invention, and FIG. 8 is a diagram for explaining the method of storing membership functions in the register, and FIG. 8 is the operation contents of the fuzzy control instruction according to the present invention. FIG. 9 is a flowchart illustrating an example of a ladder diagram type program used in a typical programmable controller. 1...CPU 2...Memory for system program 3...Memory for user program 4...Ilo 5...System bus 20...Instruction body 21...Condition input 22.
...Intermediate output register clear input 23...Execution output
24...Input register display 25...Instruction name 26...Rule storage register display 27...Output register display Agent Patent attorney Rule Chika Ken Yudo Daishimaru Ken Figure 1 2 C Figure 2 Figure 3 Figure Figure Figure Figure Figure 3

Claims (1)

[Claims] In a programmable controller that performs control by executing a sequence program programmed using a ladder diagram, it has instructions for fuzzy control that can be written in combination with the ladder diagram, and the number of fuzzy control inputs and rules executed by the instructions. A programmable controller characterized by being configured such that it can be set arbitrarily.