Fuzzy temporal control method

Claim 2

A fuzzy control method for controlling an object in accordance with control rules, the control rules having condition parts and action parts, each condition part having a fuzzy proposition representing an indefinite state of the object at an indefinite time, said method comprising the steps of: generating state measurement values of the object; calculating a degree of confidence of the fuzzy proposition of the condition part of each control rule in accordance with the state measurement values; calculating a degree of confidence of the satisfaction of the condition part of each control rule by using the degree of confidence of each said fuzzy proposition; selecting one of the control rules according to the degree of confidence of the satisfaction of each condition part; and controlling the object according to the action part of the selected control rule.

Claim 3

The fuzzy control method according to claim 1, further comprising the step of: generating reference state values of said object, and wherein said step of calculating a degree of confidence of the fuzzy proposition comprises the step of subtracting the state measurement values from the reference state values.

Claim 4

A fuzzy control method for controlling an object in accordance with control rules, the control rules having condition parts and action parts, each condition part having a fuzzy proposition representing an indefinite state of the object at an indefinite time, said method comprising the steps of: generating state measurement values at predetermined sampling times; generating reference state values for the object; subtracting the state measurement values from the reference state values thereby obtaining control error values; calculating a degree of confidence of the fuzzy proposition of the condition part of each control rule in accordance with the control error values; calculating a degree of confidence of the satisfaction of the condition part of each control rule according to the degree of confidence of each said fuzzy proposition; selecting a control rule according to the degree of confidence of the satisfaction of each said condition part; and controlling the object according to the action part of the selected control rule.

Claim 5

A fuzzy control method for controlling an object in accordance with control rules, the control rules having condition parts and action parts, each condition part having a fuzzy proposition representing an indefinite state of the object at an indefinite time, said method comprising the steps of: generating state measurement values at predetermined sampling times; generating reference state values for the object; subtracting the state measurement values from the reference state values thereby obtaining control error values; calculating a degree of confidence of the fuzzy proposition of the condition part of each control rule in accordance with the control error values; calculating a degree of confidence of the satisfaction of the condition part of each control rule according to the degree of confidence of each said fuzzy proposition; selecting a control rule according to the degree of confidence of the satisfaction of each said condition part; and controlling the object according to the action part of the selected control rule.

Claim 6

The fuzzy control method according to any one of claims 1 to 3, wherein: each said fuzzy proposition is represented by a membership function having an indefinite time as a variable.

Claim 7

The fuzzy control method according to any one of claims 1 to 3, wherein: each said fuzzy proposition is represented by a membership function having an indefinite elapsed time as a variable.

Claim 8

A fuzzy control apparatus comprising: an input for receiving a controlled variable signal from an object, said controlled variable signal being sampled at a predetermined sampling times; a control reference value signal generating circuit for generating a control reference value signal; a storage circuit for storing a plurality of fuzzy control rules having condition parts and action parts, wherein each condition part includes a fuzzy proposition, the condition part of at least one fuzzy control rule including a fuzzy proposition which represents the object in terms of an indefinite time and an indefinite state; a fuzzy control execution circuit, coupled to said input, said control reference value signal generating circuit, and said storage circuit comprising: means for determining a control error between the controlled variable signal and the control reference value signal; means for determining a degree of confidence of the fuzzy proposition of the condition part of said at least one fuzzy control rule using a predetermined function and the control error; means for evaluating the fuzzy propositions of the condition parts of the remaining fuzzy control rules; means for determining a degree of confidence of a satisfaction of the condition part of each fuzzy control rule on the basis of the determined degree of confidence of the condition parts of said at least one fuzzy control rule and the evaluated fuzzy propositions of the condition parts of the remaining fuzzy control rules; and means for selecting one of the fuzzy control rules on the basis of the degrees of confidence of the satisfaction of the condition parts; and means for outputting an operation variable to control the object based upon the action part of the selected fuzzy control rule.

Claim 9

A fuzzy control apparatus comprising: an input for receiving a controlled variable signal from an object, said controlled variable signal being sampled at a predetermined sampling times; a control reference value signal generating circuit for generating a control reference value signal; a storage circuit for storing a plurality of fuzzy control rules having condition parts and action parts, wherein each condition part includes a fuzzy proposition, the condition part of at least one fuzzy control rule including a fuzzy proposition which represents the object in terms of an indefinite time and an indefinite state; a fuzzy control execution circuit, coupled to said input, said control reference value signal generating circuit, and said storage circuit comprising: means for determining a control error between the controlled variable signal and the control reference value signal; means for determining a degree of confidence of the fuzzy proposition of the condition part of said at least one fuzzy control rule using a predetermined function and the control error; means for evaluating the fuzzy propositions of the condition parts of the remaining fuzzy control rules; means for determining a degree of confidence of a satisfaction of the condition part of each fuzzy control rule on the basis of the determined degree of confidence of the condition parts of said at least one fuzzy control rule and the evaluated fuzzy propositions of the condition parts of the remaining fuzzy control rules; and means for selecting one of the fuzzy control rules on the basis of the degrees of confidence of the satisfaction of the condition parts; and means for outputting an operation variable to control the object based upon the action part of the selected fuzzy control rule.

Claim 10

The fuzzy control apparatus of claim 6, wherein the indefinite time is represented by the indefinite time t and wherein the means for determining a degree of confidence of the fuzzy proposition of the condition part of said at least one fuzzy control rule uses the following predetermined function: max(min(f.sub.T (t),f.sub.A (e.sub.i (t.sub.c -t)))) t=0, tmax wherein f T is a membership function having time as a variable; f A is a membership function having a state of the object as a variable; tmax is the upper limit of an evaluation range of a fuzzy temporal quantifier T when present time is 0; e i is a control error at the present time; and t c is the present time.

Claim 11

The fuzzy control apparatus of claim 6, wherein when the indefinite time is represented by an indefinite elapsed time t and wherein the means for determining a degree of confidence of the fuzzy proposition of the condition part of said at least one fuzzy control rule uses the following predetermined function; ##EQU2## wherein f T is a membership function having time as a variable; f A is a membership function having a state of the object as a variable; tmax is the upper limit of an evaluation range of a fuzzy temporal quantifier T when present time is 0; e i is a control error at the present time; and t c is the present time.