Multi-region fuzzy logic control system with auxiliary variables

• US 6,041,320 A
• Filed: 08/23/1993
• Issued: 03/21/2000
• Est. Priority Date: 08/23/1993
• Status: Expired due to Term

• Current Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Original Assignee: Fisher Controls International LLC (Emerson Electric Company)
• Inventors: Qin, Si-Zhao Joe, Borders, Guy Thompson
• Primary Examiner(s): Hafiz, Tariq R.
• Application Number:
  US08/110,506
• Time in Patent Office: 2,402 Days
• Field of Search: 395/3, 395/900, 395/61, 395/906, 395/10
• US Class Current: 706/1
• CPC Class Codes: G05B 13/0275 using fuzzy logic only

Abstract

A fuzzy control system for industrial process control utilizes an auxiliary process variable to determine which of several regions of different gain a non-linear process is operating. Based upon that determination and fuzzy input signals, the fuzzy controller provides a process control signal which corresponds to that region.

Claims

1. A control system for controlling a non-linear process as a function of a process error signal, the non-linear process having a plurality of regions of operation, the control system comprising:

means for producing an auxiliary variable signals which is indicative of an instantaneous region of operation for the non-linear process, the instantaneous region of operation being one of the plurality of regions of operation; and

a fuzzy logic controller for producing a process control output signal for controlling the non-linear process in the instantaneous region of operation as a function of the process error signal and the auxiliary variable signal.

2. The control system of claim 1 further including means for producing a change in process error signal and wherein the fuzzy logic controller is configured for producing the process control output signal as a function of the process error signal, the change in process error signal and the auxiliary variable signal.

3. The control system of claim 1 wherein the fuzzy logic controller is configured to produce the process control output signal for producing proportional and integral control for the non-linear process.

4. The control system of claim 1 wherein the fuzzy logic controller is configured to produce the process control output signal that produces proportional, integral and derivative control for the non-linear process.

5. The control system of claim 1 wherein the fuzzy logic controller is configured to produce the process control output signal as a function of the process error signal and a plurality of sets of fuzzy rules for the plurality of regions of operation for the non-linear process, respectively, the fuzzy logic controller selecting one of the plurality of sets of fuzzy rules based upon the auxiliary variable signal.

6. The control system of claim 1 wherein the fuzzy logic controller includes:

fuzzification means for converting the process error signal to a fuzzy process error signal value;

fuzzy engine means for producing a fuzzy process control output signal value as a function of the fuzzy process error signal value and the auxiliary variable signal; and

defuzzification means for converting the fuzzy process control output signal value to the process control output signal.

7. The control system of claim 6 wherein the fuzzification means is configured for converting the auxiliary variable signal to a fuzzy auxiliary variable value and wherein the fuzzy engine is configured for producing the fuzzy process control output signal as a function of the fuzzy auxiliary variable value and a plurality of sets of fuzzy rules for the operating regions of the process with the fuzzy engine means selecting one of the plurality of sets of fuzzy rules based upon the fuzzy auxiliary variable value.

8. A control system for producing derivative, proportional and integral control of a process having non-linear process gain for each of a plurality of different operating regions of the process, the control system comprising:

means for producing an auxiliary value indicative of an operating region in which the process is operating wherein the operating region in which the process is operating is one of the plurality of different operating regions of the process; and

a fuzzy logic controller for producing a process control signal value which provides proportional, integral and derivative control of the process for the operating region in which the process is operating, the process control signal value being a function of a process signal error value, change in process error value, and the auxiliary value.

9. The control system of claim 8 wherein the process control signal value is produced as a function of a plurality of sets of fuzzy rules for the different operating regions of the process, the fuzzy logic controller selecting one of the plurality of sets of fuzzy rules based upon the auxiliary variable value.

10. The control system of claim 8 wherein the fuzzy logic controller includes:

fuzzification means for converting the process error signal to a fuzzy process error signal value and converting the change in process error signal to a fuzzy change in process error signal value;

fuzzy engine means for producing a fuzzy process control output signal value as a function of the fuzzy process error signal value, the fuzzy change in process error signal value and the auxiliary variable signal; and

defuzzification means for converting the fuzzy process control output signal value to the process control signal value.

11. The control system of claim 10 wherein the fuzzification means is configured for converting the auxiliary variable signal to a fuzzy auxiliary variable value and wherein the fuzzy engine is configured for producing the fuzzy process control output signal as a function of the fuzzy auxiliary variable value and a plurality of sets of fuzzy rules for the different operating regions of the process with the fuzzy engine selecting one of the plurality of sets of fuzzy rules based upon the fuzzy auxiliary variable value.

12. A control system for controlling a non-linear process as a function of a process error signal and a process change in error signal, the non-linear process having a plurality of operating regions, the control system comprising:

means for producing an auxiliary variable signal which is indicative of a region of operation in which the non-linear process is operating, wherein the region of operation is one of the plurality of operating regions; and

a fuzzy logic controller for producing a process control output signal as a function of the process error signal, the process change in error signal, and a plurality of sets of fuzzy rules for the different operating regions of the process, the fuzzy logic controller selecting one of the plurality of sets of fuzzy rules based upon the auxiliary variable signal.

13. The control system of claim 12 wherein the fuzzy logic controller is configured to produce the process output signal that provides proportional and integral control for the non-linear process.

14. The control system of claim 12 wherein the fuzzy logic controller is configured to produce the process output signal that produces proportional integral and derivative control for the non-linear process.

15. A control system for controlling a non-linear process as a function of a process error signal and a process change in error signal, the control system comprising:

means for producing an auxiliary variable signal to define an identified region of process operation from a plurality of operating regions for the non-linear process; and

a fuzzy logic controller for producing a process control output signal for controlling the non-linear process in the identified region as a function of the process error signal, the process change in error signal and the auxiliary variable signal.

16. The control system of claim 15 wherein the fuzzy logic controller is configured to produce the process control output signal for providing proportional and integral control for the non-linear process.

17. The control system of claim 15 wherein the fuzzy logic controller is configured to produce the process control output signal for producing proportional, integral and derivative control for the non-linear process.

18. The control system of claim 15 wherein the fuzzy logic controller is configured for producing the process control output signal as a function of a plurality of a plurality of sets of fuzzy rules for the different operating regions of the process, the fuzzy logic controller selecting one of the plurality of sets of fuzzy rules based upon the auxiliary variable signal.

19. A system for controlling a non-linear process comprising:

means for producing a controller parameter value indicative of an instantaneous region of process operation, the instantaneous region of process operation being one of a plurality of operating regions for the non-linear process; and

a fuzzy engine linked to the non-linear process for producing a process control output signal as a function of the process error signal, the controller parameter value and one of a plurality of sets of fuzzy rules for the different operating regions of the process.

20. A control system suited for producing a process control signal for controlling a non-linear process having first and second regions of operation, the control system comprising:

means for producing an auxiliary process signal having predetermined characteristics in each of first and second regions of process operation;

a fuzzy logic controller including;

a first fuzzy membership function relating a process error signal to the process control signal for process operation in the first region;

a second fuzzy membership function relating the process error signal to the process control signal for process operation in the second region; and

a fuzzy inference engine for applying one of the first and second fuzzy membership functions selected based on the auxiliary process signal to a process error signal value for inferring a process control signal value.

21. The control system of claim 20 wherein the first fuzzy membership function relates the process error signal to the process control signal to produce proportional and integral control for the first region of process operation and the second fuzzy membership function relates the process error signal to the process control signal to produce proportional and integral control for the second region of process operation.

22. The control system of claim 20 and further including signal combining means for combining a setpoint value and a process output signal value to produce a process error signal value and a change in process error value.

23. The control system of claim 20 wherein the auxiliary process signal is the process control signal value.

24. The control system of claim 20 wherein the auxiliary process signal is a process output signal value.

25. The control system of claim 20 wherein the first fuzzy membership function as well as the second fuzzy membership function are fuzzy rules that if combined with fuzzy variables are evaluated by fuzzy logic to provide fuzzy inferences.

26. The control system of claim 20 wherein the first fuzzy membership function rules which are identical to the second fuzzy membership function rules are combined forming a single membership function.

27. The control system of claim 20 wherein the fuzzy inference engine selects one of the first and second membership functions based on the auxiliary process signal.

28. The control system of claim 20 and further including means for selecting one of the first and second membership functions based on the auxiliary process signal.

29. A control system for controlling a process having non-linear process gain, the process having first and second regions of process gain, the control system comprising:

a circuit, connected between a process and a fuzzy logic controller, for producing an auxiliary process variable value that is indicative of first and second regions of process gain;

the fuzzy logic controller for producing a control signal value to the process, the fuzzy logic controller including;

a first fuzzy membership function that is selected to produce proportional and integral control suitable for the first region of process gain;

a second fuzzy membership function that is selected to produce proportional and integral control suitable for the second region of process gain; and

a fuzzy engine for applying one of the first and second fuzzy membership functions selected based on an indicated region of process gain determined from the auxiliary process variable value to a process error signal value and a change in process error value to infer the control signal value for maintaining proper process operation in the first and second regions of process gain.

30. A method for controlling a non-linear process as a function of a process error signal, the method comprising:

determining a region of process operation of the non-linear process from a plurality of operating regions based on an auxiliary process variable signal; and

applying a selected set of fuzzy rules from a plurality of sets of fuzzy rules for the plurality of operating regions of the process to the process error signal for inferring a process control output signal, the selected set of fuzzy rules applied being based upon the operating region of the process determined from the auxiliary process variable signal.

31. The method for controlling a non-linear process of claim 30 and further including converting the process error signal to a fuzzy process error signal value and wherein applying the selected set of fuzzy rules from a plurality of sets of fuzzy rules includes selecting a fuzzy rule from the selected set of fuzzy rules based on the fuzzy process error signal for inferring a fuzzy process control signal value.

32. The method for controlling a non-linear process of claim 31 further including converting the fuzzy process control signal value to the process control output signal.

33. The method for controlling a non-linear process of claim 31 wherein selecting a fuzzy rule from the selected set of fuzzy rules is performed utilizing a fuzzy logic controller.