Fuzzy controller for selecting an input signal

Abstract

In a fuzzy controller comprising a fuzzy inference device for subjecting one or a plurality of input signals to fuzzy inference operations using membership functions in accordance with a predetermined rule and outputting a non-fuzzy value determined on the basis of the results of the inferences, there are provided an input signal level converter for converting signal levels of the input signals on the input side of the fuzzy inference device and an output signal level converter for converting a signal level of the output signal on the output side of the fuzzy inference device.

What is claimed is: 
     
       1. A fuzzy controller comprising: fuzzy inference means for subjecting one or a plurality of input signals to fuzzy inference operations using membership functions in accordance with a predetermined rule and outputting a non-fuzzy value determined on the basis of the results of the inference;   scale factor control means for selecting an input signal whose absolute value is a maximum value of said input signals and for respectively outputting a multiplier value and a divider value which correspond to said maximum absolute value;   input-side multiplication means for multiplying the value of said input signals by said multiplier value obtained from said scale factor control means; and   output-side division means for dividing the value of said output signal outputted from said fuzzy inference means by said divider value obtained from said scale factor control means.   
     
     
       2. A fuzzy .[.logic.]. controller for controlling a controllable system, said fuzzy .[.logic.]. controller comprising: a plurality of input terminals for receiving respective applied .[.first.]. .Iadd.plurality of .Iaddend.input signals;   a fuzzy inference portion for subjecting one or a plurality of .[.second.]. .Iadd.fuzzy inference .Iaddend.input signals to fuzzy inference operations using membership functions in accordance with a predetermined rule and outputting a non-fuzzy .[.signal.]. .Iadd.value .Iaddend.determined on the basis of the results of the interference operation;   a scale factor control section for selecting one of said applied .[.first.]. .Iadd.plurality of .Iaddend.input signals whose absolute value is a maximum value of said applied .[.first.]. .Iadd.plurality of .Iaddend.input signals and for outputting a multiplier value corresponding to said maximum absolute value;   an input-side multiplier for multiplying said applied .[.first.]. .Iadd.plurality of .Iaddend.input signals by said multiplier value obtained from said scale factor control section to produce said .[.second.]. .Iadd.fuzzy inference .Iaddend.input signals which are applied to said fuzzy inference portion; and   a controllable system connected to receive and be controlled by said non-fuzzy .[.signal.]. .Iadd.value.Iaddend..   
     
     
       3. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 2, further comprising output-side division means for dividing .Iadd.a signal representing .Iaddend.the value of said non-fuzzy output .[.signal.]. .Iadd.value .Iaddend.outputted from said fuzzy inference means by a divider value. 
     
     
       4. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 3, wherein said divider value is obtained from said scale factor control means which determines said divider value in correspondence with said maximum absolute value. 
     
     
       5. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claims 1 or 2, wherein said scale factor control means.Iadd., or scale factor control section, .Iaddend.further comprises a MAX circuit for selecting .[.an.]. .Iadd.a .Iaddend.signal value whose absolute value is a maximum of said plurality .Iadd.of .Iaddend.input signals. 
     
     
       6. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 5, wherein said MAX circuit comprises: input signal receiving means, for receiving .Iadd.input .Iaddend.signals I 1  -I n  ; and   maximum value calculating means respectively connected to said input .Iadd.signal .Iaddend.receiving means for providing an output signal whose absolute value is the maximum value of said input signals I 1  -I n .   
     
     
       7. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 6, .[.where.]. .Iadd.wherein .Iaddend.said scale factor control section .Iadd.or means .Iaddend.comprises: a comparator for comparing said absolute value to low and high voltage reference values; and   a scale factor determining portion responsive to said comparator for providing a multiplier value and a divider value.   
     
     
       8. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 7, wherein when said input signals I 1  -I n  fall in a steady state operation range of said controller, said multiplier and divider values produced from said scale factor control means .Iadd.or section .Iaddend.are increased to exceed one (1). 
     
     
       9. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 8, wherein result of said operations in said fuzzy inference means.Iadd., portion, .Iaddend.are provided as an antecedent to said scale factor control means .Iadd.or section .Iaddend.as input signals. 
     
     
       10. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.according to claim 7, wherein when said signals I 1  -I n  fall in a greatly changing operational state of said controller, said multiplier and divider values are decreased by said scale factor control means .Iadd.or section .Iaddend.to a value approaching one (1). 
     
     
       11. The .[.apparatus.]. .Iadd.fuzzy controller .Iaddend.of claim 10, wherein .[.said.]. .Iadd.a .Iaddend.signal .Iadd.determined by and .Iaddend.outputted from said .[.input side.]. .Iadd.input-side .Iaddend.multiplication means is limited within a predetermined range. 
     
     
       12. A method of fully control, comprising the steps of: selecting one input signal whose absolute value is a maximum value out of a plurality if input signals;   determining a multiplier value corresponding to said maximum absolute value;   multiplying said plurality of input signals by said multiplier value; and   subjecting said plurality of multiplied input signals to fuzzy inference operations using membership functions in accordance with predetermined rules. .Iadd.13. A fuzzy controller for controlling a controllable system, said fuzzy controller comprising:   a plurality of inputs for receiving input signals;   a first signal level adjusting circuit connected to said plurality of inputs for adjusting the signal levels of at least one input signal applied to said inputs in accordance with at least one first control signal;   a fuzzy inference circuit responsive to the input signals passing through said first signal level adjusting circuit for subjecting at least one of said input signals passing through said level adjusting circuit to fuzzy inference operations using at least one membership function in accordance with at least one predetermined rule, and outputting at least one non-fuzzy value determined on the basis of the results of the inference operations;   second signal level adjusting circuit coupled to said fuzzy inference circuit for receiving and adjusting the level of at least one of said at least one non-fuzzy value in accordance with at least one second control signal; and   a control circuit for supplying at least one of said first and second control signals to said first and second signal level adjusting circuits to thereby control the adjustment level of said first and second signal level adjusting circuits, said control circuit providing said first and second control signals in response to at least one third control signal   
     
     
        applied thereto. .Iaddend..Iadd.14.  A fuzzy controller as in claim 13, wherein said control circuit provides said first and second control signals based at least in part on a signal characteristic of said input signals. .Iaddend..Iadd.15. A fuzzy controller as in claim 13, wherein said control circuit provides said first and second control signals based at least in part on a signal characteristic of said at least one non-fuzzy value. .Iaddend..Iadd.16. A fuzzy controller as in claim 15, wherein said control circuit provides said first and second control signals based at least in part on a signal characteristic of said input signals and on a signal characteristic of said at least one non-fuzzy value. .Iaddend..Iadd.17. A fuzzy controller as in claim 13, wherein said control circuit provides said first and second control signals based at least in part on a signal characteristic of said input signals and on a signal characteristic of said at least one non-fuzzy value. .Iaddend..Iadd.18. A fuzzy controller as in claim 13, wherein said control circuit provides said first and second control signals based at least in part on an output signal from a system which is controlled by said at least one non-fuzzy value. .Iaddend..Iadd.19. The fuzzy controller of claim 13, wherein said first signal level adjusting circuit changes the gain and offset of at least one of said input signals in response to said at least one first control signal. .Iaddend..Iadd.20. The fuzzy controller of claim 19, wherein said first signal level adjusting circuit includes a multiplier circuit for adjusting a gain value and an adder circuit for receiving the output of said multiplier circuit and for adjusting an offset value. 
     
     
        .Iaddend..Iadd.21.  The fuzzy controller of claim 13, wherein said second signal level adjusting circuit changes the gain and offset of said at least one non-fuzzy value in response to said at least one second control signal. .Iaddend..Iadd.22. The fuzzy controller of claim 21, wherein said first signal level adjusting circuit includes a multiplier circuit for adjusting a gain value and an adder circuit for receiving the output of said multiplier circuit and for adjusting an offset value. .Iaddend..Iadd.23. The fuzzy controller as in claim 13, wherein said control circuit includes a CPU and a memory. .Iaddend.