JPH0572613B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a fuzzy inference device for fuzzy adjustment performed by applying so-called "fuzzy logic".

[Conventional technology]

In the inference calculation in the fuzzy inference device,
When performing fuzzy inference using fuzzy variables expressed by membership functions in the antecedent and consequent propositions of a control rule, the composition rule of inference is generally often used. That is, when a plurality of control rules are given, the ambiguous relationship R is determined in advance, and the membership function for the output is determined using the input value expressed by the membership function.

FIG. 2 is an explanatory diagram showing an example of a calculation method that simplifies such conventional fuzzy inference. below,
See Figure 2.

Suppose that there is a control rule expressed by the following equation.

IF x01=A11, x02=A21, x03=A31,...
x0k=Ak1 THEN u=B (If x01=A11, x02=A21, x03=A31,
. . . If x0k=Ak1, the manipulated variable u is B) The blocks surrounded by broken lines in FIG. In the same figure, A11, A21, A31, ...Ak1 are the membership functions of the antecedent part, B1, ...BN are the membership functions of the consequent part, respectively,
x01, x02, x03, ...x0k are each measured values.

1a, 1b,...1k are arithmetic circuits (indicated by the symbol F) that calculate membership function values, 2 is a minimum value arithmetic circuit (indicated by the symbol Λ) that takes the minimum value from multiple inputs and outputs it, and 3 is an input. The membership function B1 to be calculated is multiplied by a constant (minimum value calculation circuit 2
(output times)) and outputs the circuit (indicated by *), 4
is a maximum value calculation circuit (indicated by symbol V) that always selects and outputs the maximum value from among multiple inputs.

Next, the control operation will be explained. In the arithmetic circuit 1a,
The membership function value is determined from the membership function A11 of the antecedent part and the measured value x01 and output. The same applies to the other arithmetic circuits 1b, . . . 1k. The minimum value calculation circuit 2 receives each output from the calculation circuits 1a, 1b, . . . 1k, selects the minimum value from among them, and outputs it as the effectiveness α1. The antecedent part calculation ends when this validity α1 is determined.

Next, we move on to the consequent operation. The constant multiplier circuit 3 multiplies the consequent membership function B1 by the validity α1, outputs the result as B1', and sends it to the maximum value calculation circuit 4. The maximum value calculation circuit 4 receives the results obtained by similarly calculating other control rules as B2'...BN'. In the maximum value calculation circuit 4, those inputs
The maximum value is always selected from among the waveforms of B1'...BN' and output as the membership function B for output. To actually obtain the operation output value from the output membership function B, use the membership function B
A method is adopted in which the operation output value is obtained by finding the center of gravity in the area formed by the function waveform and the horizontal axis.

By the way, there are a finite number of fuzzy variables (B1, ...BN in Figure 2) in the consequent part of the control rule,
B1 is NB (negative and large), B2 is NM (negative and medium)
, B3 is NS (negative and small), B4 is ZE (as is), B5 is PS (positive and small), B6 is PM (positive and medium), B7 is PB (positive and large). , respectively, and the fuzzy variables in the consequent part are these 7
It shall consist of stages.

However, the number of control rules is not limited to seven, and there may be many. However, even in that case, the fuzzy variables in the consequent of each control rule are the seven mentioned above: NB (large negative), NM (medium negative), NS (small negative), ZE (as is), PS
(positive and small), PM (positive and medium), and PB (positive and large).

Therefore, it is determined in advance which control rule corresponds to which consequent fuzzy variable and the correspondence is established.

[Problem that the invention seeks to solve]

In the conventional fuzzy inference device described above, the control rules used are all equivalent, and the effectiveness determined in the antecedent part calculation is used as is in the consequent part calculation to perform the inference calculation.

Therefore, even if it was desired to assign a level to each control rule for various reasons, there was no way to do so.

Therefore, the present invention provides a fuzzy inference device that includes:
The problem to be solved is to be able to assign levels to the control rules used.

[Means for solving problems]

The present invention is a device that performs fuzzy inference calculations based on measured value input and creates an operation output for a controlled object, and is comprised of the above-mentioned measured value input device, storage device, arithmetic device, and output device.

The storage device includes a control rule antecedent portion storage device that stores an antecedent portion of a control rule corresponding to the measurement value input, and a “weight” (a numerical value between 0.0 and 1.0) determined for each control rule antecedent portion. The control rule weight storage device stores a control rule weight storage device that stores a control rule (for which the control rules are divided into levels according to the magnitude of the value), and a control rule consequent storage device that stores a control rule consequent portion.

The arithmetic device calculates the effectiveness of the antecedent proposition (the grade of the antecedent proposition) using the control rule given from the corresponding control rule antecedent storage device and the measured value given from the corresponding input device. a control rule antecedent part calculator that calculates the minimum value of , etc.), and a multiplication unit that multiplies the effectiveness from the control rule antecedent part calculator by the corresponding weight of the control rule from the weight storage device. A control rule consequent part that performs a consequent calculation using a vector, the effectiveness weighted in the multiplier, and a membership function representing a control rule consequent proposition from the control rule consequent part storage device. including a computing unit.

The output device includes a manipulated value output device that outputs the consequent calculation result to the controlled object.

[Effect]

In the fuzzy inference device configured as described above, the effectiveness of the antecedent proposition (defined as the minimum value of the grade of the antecedent proposition) is determined, and the effectiveness is determined based on the effectiveness given for each control rule. The effectiveness can be converted by multiplying by a weight (a numerical value between 0.0 and 1.0 that is used to classify the control rule into levels). Further, the calculation of the consequent part is performed using a membership function representing its effectiveness and the consequent proposition of each control rule.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing one embodiment of the present invention. In this embodiment, there are n control rules, and an antecedent part calculation is performed for each of them.

In the figure, 11a and 11n are control rule weight storage devices, 12a and 12n are control rule antecedent storage devices, 13a and 13n are measurement value input devices, and 14a and 14n are control rule antecedent storage devices, respectively. 15a and 15n are multipliers, 16 is a control rule consequent storage device,
17 is a control rule consequent calculator, and 18 is a manipulated value output device.

The control rule antecedent part calculator 14a calculates the effectiveness α1 using a certain control rule 1 given from the control rule antecedent part storage device 12a and the measured value 1 given from the measured value input device 13a. Output.
In the multiplier 15a, the control rule antecedent part calculator 14a
The effectiveness degree α1 from 1 is multiplied by the weight of control rule 1 retrieved from the weight storage device 11a and output. Similarly, another multiplier 15n multiplies the effectiveness αn obtained by the control rule antecedent calculator 14n by the weight of the control rule n retrieved from the weight storage device 11n and outputs the result.

The control rule consequent part calculator 17 receives the weighted effectiveness of each multiplier, and
This and the control rule 1 retrieved from the storage device 16
The consequent part calculation is performed using the consequent part of ~n. The result is output to the manipulated value output device 18.

〔Effect of the invention〕

As explained above, according to the present invention, each control rule used can be divided into levels in inference calculations in a fuzzy inference device, so it is possible to simultaneously process independent and mutually exclusive events by setting the level of control rules. There are advantages.

Another advantage is that once a control rule has been input, it can be selected to be used or not depending on the input value situation. That is, when not in use, the weight may be set to zero.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention;
FIG. 2 is an explanatory diagram showing an example of a conventional fuzzy inference device. Explanation of symbols, 11a, 11n...Control rule weight storage device, 12a, 12n...Control rule antecedent part storage device, 13a, 13n...Measurement value input device, 14a, 14n...Control rule antecedent unit,
15a, 15n...multiplier, 16...control rule consequent part storage device, 17...control rule consequent part arithmetic unit,
18... Manipulated value output device.

Claims (1)

[Scope of Claims] 1. A device that performs fuzzy inference calculations based on measured value input from a controlled object and creates an operation output for the controlled object, comprising an input device, a storage device, a calculation device, and an output device for the measured values. The storage device includes a control rule antecedent portion storage device that stores an antecedent portion of a control rule in response to a measurement value input, and a “weight” determined for each of the control rule antecedent portions. The arithmetic unit includes a weight storage device for storing the consequent part of the control rule, and a control rule consequent part storage device for storing the consequent part of the control rule, and the arithmetic unit is configured to calculate the control rule given from the corresponding control rule antecedent part storage device, and the control rule consequent part storage device to store the consequent part of the control rule. a control rule antecedent part calculator that calculates the effectiveness using the measurement value given from the above input device; a multiplier that multiplies the weight of the control rule; a weighted effectiveness in the multiplier;
a control rule consequent part calculator that performs a consequent part calculation using the control rule consequent part from the control rule consequent part storage device; A fuzzy inference device comprising: a manipulated value output device that outputs an output to a controlled object.