Fuzzy logic design generator using a neural network to generate fuzzy logic rules and membership functions for use in intelligent systems

• US 5,579,439 A
• Filed: 03/24/1993
• Issued: 11/26/1996
• Est. Priority Date: 03/24/1993
• Status: Expired due to Fees

• Current Assignee: National Semiconductor Corporation (Texas Instruments, Inc.)
• Original Assignee: National Semiconductor Corporation (Texas Instruments, Inc.)
• Inventors: Khan, Emdadur R.
• Primary Examiner(s): Downs, Robert W.
• Application Number:
  US08/036,634
• Time in Patent Office: 1,343 Days
• Field of Search: 395/3, 395/900, 395/22, 395/11
• US Class Current: 706/2
• CPC Class Codes: G06N 3/043 based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
• CPC Class Codes: G06N 7/046 Implementation by means of a neural network neural networks using fuzzy logic G06N3/043
• CPC Class Codes: Y10S 706/90 Fuzzy logic

Abstract

A fuzzy logic design generator for providing a fuzzy logic design for an intelligent controller in a plant control system includes an artificial neural network for generating fuzzy logic rules and membership functions data. These fuzzy logic rules and membership functions data can be stored for use in a fuzzy logic system for neural network based fuzzy antecedent processing, rule evaluation and defuzzification, thereby avoiding heuristics associated with conventional fuzzy logic algorithms. The neural network, used as a fuzzy rule generator to generate fuzzy logic rules and membership functions for the system'"'"'s plant controller, is a multilayered feed-forward neural network based upon a modified version of a back-propagation neural network and learns the system behavior in accordance with input and output data and then maps the acquired knowledge into a new non-heuristic fuzzy logic system. Interlayer weights of the neural network are mapped into fuzzy logic rules and membership functions. Antecedent processing is performed according to a weighted product of the antecedents. One layer of the neural network is used for performing rule evaluation and defuzzification.

Claims

1. An artificial neural network for generating pluralities of signals representing a plurality of fuzzy logic rules and a plurality of fuzzy logic membership functions, comprising:

first neural means for receiving a plurality of signals representing input data for an intelligent system and for providing fuzzified data which corresponds to said input data;

second neural means coupled to said first neural means for receiving said fuzzified data and in accordance therewith generating a plurality of membership signals which correspond to a plurality of fuzzy logic membership functions; and

third neural means coupled to said second neural means for receiving said plurality of membership signals and in accordance therewith generating a plurality of intermediate signals and a plurality of logic rule signals which represent a plurality of fuzzy logic rules and in accordance therewith generating an output signal which represents defuzzified data, wherein said third neural means includes a single output neuron for receiving and processing said plurality of intermediate signals and in accordance therewith generating said output signal;

wherein said first, second and third neural means cooperate together by performing a learning process with back-propagation of an output error signal based upon said output signal, and wherein said output error signal is propagated serially back from an output of said third neural means through and successively processed by said third, second and first neural means.

2. An artificial neural network as recited in claim 1, wherein said first neural means comprises a first plurality of artificial neurons for fuzzifying said input data to produce said fuzzified data.

3. An artificial neural network as recited in claim 1, wherein said second neural means comprises a second plurality of artificial neurons for generating a plurality of neuron output signals as said plurality of membership signals.

4. An artificial neural network as recited in claim 1, wherein said third neural means multiplies selected ones of said received plurality of membership signals.

5. An artificial neural network as recited in claim 1, wherein said third neural means comprises a third plurality of artificial neurons which includes a plurality of output weights corresponding to said plurality of logic rule signals.

6. An artificial neural network as recited in claim 5, wherein said third plurality of artificial neurons multiply selected ones of said received plurality of membership signals.

7. An artificial neural network as recited in claim 1, further comprising comparison means coupled to said third neural means for receiving therefrom an output signal representing a plant control signal, for receiving an input signal representing a desired control signal, and for comparing said input desired control signal with said output plant control signal and in accordance therewith generating said output error signal.

8. An artificial neural network as recited in claim 7, wherein said comparison means comprises a comparator.

9. An artificial neural network as recited in claim 1, further comprising fourth neural means coupled to said third neural..means for receiving and summing a plurality of signals therefrom to provide an output sum signal representing defuzzified data.

10. An artificial neural network as recited in claim 9, wherein said fourth neural means comprises an artificial neuron.

11. An artificial neural network as recited in claim 9, further comprising comparison means coupled to said fourth neural means for receiving therefrom said output sum signal, for receiving an input signal representing a desired control signal, and for comparing said input desired control signal with said output sum signal and in accordance therewith generating said output error signal, wherein said output sum signal represents a plant control signal.

12. An artificial neural network as recited in claim 11, wherein said comparison means comprises a comparator.

13. An artificial neural network for generating pluralities of signals representing a plurality of fuzzy logic rules and a plurality of fuzzy logic membership functions, comprising:

a first plurality of artificial neurons which receive a plurality of signals representing input data for an intelligent system and provide fuzzified data which correspond to said input data;

a second plurality of artificial neurons, coupled to said first plurality of artificial neurons, which receive said fuzzified data and in accordance therewith generate a plurality of membership signals which correspond to a plurality of fuzzy logic membership functions; and

a third plurality of artificial neurons, coupled to said second plurality of artificial neurons, which receive said plurality of membership signals and in accordance therewith generate a plurality of intermediate signals and a plurality of logic rule signals which represent a plurality of fuzzy logic rules and in accordance therewith generate an output signal which represents defuzzified data, wherein said third plurality of artificial neurons includes a plurality of output weights corresponding to said plurality of logic rule signals and a single output neuron which receives and processes said plurality of intermediate signals and in accordance therewith generates said output signal;

wherein said first, second and third pluralities of artificial neurons cooperate together by performing a learning process with back-propagation of an output error signal based upon said output signal, and wherein said output error signal is propagated serially back from an output of said third plurality of artificial neurons through and successively processed by said third, second and first pluralities of artificial neurons.

14. An artificial neural network as recited in claim 13, wherein said third plurality of artificial neurons multiply selected ones of said received plurality of membership signals.

15. An artificial neural network as recited in claim 13, further comprising a comparator, coupled to said third plurality of artificial neurons, which receives therefrom an output signal representing a plant control signal, receives an input signal representing a desired control signal, and compares said input desired control signal with said output plant control signal and in accordance therewith generates said output error signal.

16. An artificial neural network as recited in claim 13, further comprising an artificial neuron, coupled to said third plurality of artificial neurons, which receives and sums a plurality of signals therefrom to provide an output sum signal representing defuzzified data.

17. An artificial neural network as recited in claim 16, further comprising a comparator, coupled to said artificial neuron, which receives therefrom said output sum signal, receives an input signal representing a desired control signal, and compares said input desired control signal with said output sum signal and in accordance therewith generates said output error signal, wherein said output sum signal represents a plant control signal.

18. A computer-implemented method for generating pluralities of signals representing a plurality of fuzzy logic rules and a plurality of fuzzy logic membership functions, comprising the computer-implemented steps of:

receiving a plurality of signals representing input data for an intelligent system;

generating a plurality of fuzzified data signals representing fuzzified data which corresponds to said input data;

generating in accordance with said plurality of fuzzified data signals a plurality of membership signals which correspond to a plurality of fuzzy logic membership functions;

generating in accordance with said plurality of membership signals a plurality of intermediate signals and a plurality of logic rule signals which represent a plurality of fuzzy logic rules;

processing said plurality of intermediate signals with a single output neuron and in accordance therewith generating an output signal which represents defuzzified data; and

performing a learning process with back-propagation of an output error signal based upon said output signal in cooperation with said steps of generating said pluralities of fuzzified data signals, membership signals and logic rule signals, wherein said back-propagation of said output error signal is performed following a completion of one iteration of said step of generating said plurality of logic rule signals and is performed successively and prior to subsequent iterations of said steps of generating said pluralities of fuzzified data signals, membership signals and logic rule signals.

19. A method as recited in claim 18, wherein said step of generating a plurality of fuzzified data signals representing fuzzified data which corresponds to said input data comprises fuzzifying said input data with a first plurality of artificial neurons to produce said plurality of fuzzified data signals.

20. A method as recited in claim 18, wherein said step of generating in accordance with said plurality of fuzzified data signals a plurality of membership signals which correspond to a plurality of fuzzy logic membership functions comprises inputting said plurality of fuzzified data signals to a second plurality of artificial neurons and outputting therefrom a plurality of neuron output signals as said plurality of membership signals.

21. A method as recited in claim 18, wherein said step of generating in accordance with said plurality of membership signals a plurality of logic rule signals which represent a plurality of fuzzy logic rules comprises outputting a plurality of weight signals corresponding to a plurality of output weights associated with a third plurality of artificial neurons, wherein said plurality of weight signals represent said plurality of logic rule signals.

22. A method as recited in claim 21, wherein said step of generating in accordance with said plurality of membership signals a plurality of logic rule signals which represent a plurality of fuzzy logic rules further comprises multiplying selected ones of said received plurality of membership signals.

23. A method as recited in claim 18, wherein said step of generating in accordance with said plurality of membership signals a plurality of logic rule signals which represent a plurality of fuzzy logic rules comprises multiplying selected ones of said received plurality of membership signals.

24. A method as recited in claim 18, further comprising the steps of:

providing an output signal representing a plant control signal;

receiving an input signal representing a desired control signal; and

comparing said input desired control signal with said output plant control signal and in accordance therewith generating said output error signal.

25. A method as recited in claim 18, further comprising the steps of:

generating a plurality of output signals associated with said plurality of logic rule signals; and

summing said plurality of output signals associated with said plurality of logic rule signals to provide an output sum signal representing defuzzified data.

26. A method as recited in claim 25, wherein said step of summing said plurality of output signals associated with said plurality of logic rule signals to provide an output sum signal representing defuzzified data comprises summing said plurality of output signals with an artificial neuron.

27. A method as recited in claim 25, further comprising the steps of:

receiving said output sum signal, wherein said output sum signal represents a plant control signal;

receiving an input signal representing a desired control signal; and

comparing said input desired control signal with said output sum signal and in accordance therewith generating said output error signal.

28. An artificial neural network for defuzzifying a plurality of fuzzy data generated in accordance with a plurality of fuzzy logic rules and membership functions, comprising:

rule-based neural means for receiving and processing a plurality of fuzzy signals representing fuzzified data which has been fuzzified in accordance with a plurality of fuzzy logic membership functions to provide in accordance therewith a plurality of processed fuzzy signals;

output weights means coupled to said rule-based neural means for receiving and weighting said plurality of processed fuzzy signals; and

single output neuron means coupled to said output weights means for receiving and summing said plurality of weighted, processed fuzzy signals to provide an output signal representing defuzzified data.

29. An artificial neural network as recited in claim 28, wherein said rule-based neural means multiplies selected ones of said plurality of fuzzy signals.

30. An artificial neural network as recited in claim 28, wherein said rule-based neural means comprises a first plurality of artificial neurons.

31. An artificial neural network as recited in claim 30, wherein said first plurality of artificial neurons multiply selected ones of said plurality of fuzzy signals.

32. An artificial neural network as recited in claim 28, wherein said plurality of fuzzy signals corresponds to a plurality of fuzzy logic membership functions.

33. An artificial neural network as recited in claim 28, wherein said output weights means comprises a plurality of interneuron weights.

34. An artificial neural network as recited in claim 33, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

35. An artificial neural network as recited in claim 28, wherein said single output neuron means comprises a single artificial neuron.

36. An artificial neural network as recited in claim 28, further comprising fuzzifier neural means coupled to said rule-based neural means for receiving and fuzzifying a plurality of data signals to provide said plurality of fuzzy signals thereto.

37. An artificial neural network as recited in claim 36, wherein said fuzzifier neural means comprises a second plurality of artificial neurons.

38. An artificial neural network for defuzzifying a plurality of fuzzy data generated in accordance with a plurality of fuzzy logic rules and membership functions, comprising:

a first plurality of artificial neurons which receive and process a plurality of fuzzy signals representing fuzzified data which has been fuzzified in accordance with a plurality of fuzzy logic membership functions to provide in accordance therewith a plurality of processed fuzzy signals;

a plurality of interneuron weights, coupled to said first plurality of artificial neurons, which receive and weight said plurality of processed fuzzy signals; and

a single artificial neuron, coupled to said plurality of interneuron weights, which receives and sums said plurality of weighted, processed fuzzy signals to provide an output signal representing defuzzified data.

39. An artificial neural network as recited in claim 38, wherein said first plurality of artificial neurons multiply selected ones of said plurality of fuzzy signals.

40. An artificial neural network as recited in claim 38, wherein said plurality of fuzzy signals corresponds to a plurality of fuzzy logic membership functions.

41. An artificial neural network as recited in claim 38, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

42. An artificial neural network as recited in claim 38, further comprising a second plurality of artificial neurons, coupled to said first plurality of artificial neurons, which receive and fuzzify a plurality of data signals to provide said plurality of fuzzy signals thereto.

43. A computer-implemented method for defuzzifying a plurality of fuzzy data generated in accordance with a plurality of fuzzy logic rules and membership functions, comprising the computer-implemented steps of:

receiving a plurality of fuzzy signals representing fuzzified data which has been fuzzified in accordance with a plurality of fuzzy logic membership functions;

processing said plurality of fuzzy signals to provide a plurality of processed fuzzy signals;

weighting said plurality of processed fuzzy signals; and

summing said plurality of weighted, processed fuzzy signals, in accordance with a single summing function, to provide an output signal representing defuzzified data.

44. A method as recited in claim 43, wherein said step of processing said plurality of fuzzy signals to provide a plurality of processed fuzzy signals comprises multiplying selected ones of said plurality of fuzzy signals.

45. A method as recited in claim 43, wherein said step of processing said plurality of fuzzy signals to provide a plurality of processed fuzzy signals comprises processing said plurality of fuzzy signals with a plurality of artificial neurons.

46. A method as recited in claim 45, wherein said step of processing said plurality of fuzzy signals to provide a plurality of processed fuzzy signals further comprises multiplying selected ones of said plurality of fuzzy signals.

47. A method as recited in claim 43, wherein said plurality of fuzzy signals corresponds to a plurality of fuzzy logic membership functions.

48. A method as recited in claim 43, wherein said step of weighting said plurality of processed fuzzy signals comprises weighting said plurality of processed fuzzy signals with a plurality of interneuron weights.

49. A method as recited in claim 48, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

50. A method as recited in claim 43, wherein said step of summing said plurality of weighted, processed fuzzy signals to provide a defuzzified output signal comprises summing said plurality of weighted, processed fuzzy signals with an artificial neuron.

51. A method as recited in claim 43, further comprising the step of receiving and fuzzifying a plurality of data signals to provide said plurality of fuzzy signals.

52. A method as recited in claim 51, wherein said step of receiving and fuzzifying a plurality of data signals to provide said plurality of fuzzy signals comprises receiving and fuzzifying said plurality of data signals with a second plurality of artificial neurons.

53. An artificial neural network for processing a plurality of input signals as a plurality of fuzzy logic rule antecedents, comprising:

input neural means for receiving and fuzzifying a plurality of input data signals to provide a plurality of fuzzified data signals corresponding to a plurality of fuzzy logic rule antecedents; and

output neural means coupled to said input neural means for receiving and multiplying together selected ones of said plurality of fuzzified data signals to provide a plurality of product signals corresponding to a plurality of fuzzy logic rule consequents.

54. An artificial neural network as recited in claim 53, wherein said input neural means comprises a first plurality of artificial neurons.

55. An artificial neural network as recited in claim 53, wherein said output neural means comprises a second plurality of artificial neurons.

56. An artificial neural network as recited in claim 53, further comprising output weights means coupled to said output neural means for receiving and weighting said plurality of product signals to provide a plurality of weighted product signals.

57. An artificial neural network as recited in claim 56, wherein said output weights means comprises a plurality of interneuron weights.

58. An artificial neural network as recited in claim 57, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

59. An artificial neural network as recited in claim 56, further comprising neural summing means coupled to said output weights means for receiving and summing said plurality of weighted product signals.

60. An artificial neural network as recited in claim 59, wherein said neural summing means comprises an artificial neuron.

61. An artificial neural network for processing a plurality of input signals as a plurality of fuzzy logic rule antecedents, comprising:

a first plurality of artificial neurons which receive and fuzzify a plurality of input data signals to provide a plurality of fuzzified data signals corresponding to a plurality of fuzzy logic rule antecedents; and

a second plurality of artificial neurons, coupled to said first plurality of artificial neurons, which receive and multiply together selected ones of said plurality of fuzzified data signals to provide a plurality of product signals corresponding to a plurality of fuzzy logic rule consequents.

62. An artificial neural network as recited in claim 61, further comprising a plurality of interneuron weights, coupled to said second plurality of artificial neurons, which receive and weight said plurality of product signals to provide a plurality of weighted product signals.

63. An artificial neural network as recited in claim 62, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

64. An artificial neural network as recited in claim 62, further comprising an artificial neuron, coupled to said plurality of interneuron weights, which receives and sums said plurality of weighted product signals.

65. A computer-implemented method for processing a plurality of input signals as a plurality of fuzzy logic rule antecedents, comprising the computer-implemented steps of:

receiving a plurality of input data signals;

fuzzifying said plurality of input data signals to provide a plurality of fuzzified data signals corresponding to a plurality of fuzzy logic rule antecedents; and

multiplying together selected ones of said plurality of fuzzified data signals to provide a plurality of product signals corresponding to a plurality of fuzzy logic rule consequents.

66. A method as recited in claim 65, wherein said step of fuzzifying said plurality of input data signals to provide a plurality of fuzzified data signals corresponding to a plurality of fuzzy logic rule antecedents comprises fuzzifying said plurality of input data signals with a first plurality of artificial neurons.

67. A method as recited in claim 65, wherein said step of multiplying together selected ones of said plurality of fuzzified data signals to provide a plurality of product signals comprises multiplying together selected ones of said plurality of fuzzified data signals with a second plurality of artificial neurons.

68. A method as recited in claim 65, further comprising the step of receiving and weighting said plurality of product signals to provide a plurality of weighted product signals.

69. A method as recited in claim 68, wherein said step of receiving and weighting said plurality of product signals to provide a plurality of weighted product signals comprises receiving and weighting said plurality of product signals with a plurality of interneuron weights.

70. A method as recited in claim 69, wherein said plurality of interneuron weights corresponds to a plurality of fuzzy logic rules.

71. A method as recited in claim 68, further comprising the step of summing said plurality of weighted product signals.

72. A method as recited in claim 71, wherein said step of summing said plurality of weighted product signals comprises summing said plurality of weighted product signals with an artificial neuron.

73. An artificial neural network for processing a plurality of input data in accordance with fuzzy logic, comprising:

input neural means for receiving and fuzzifying a plurality of input signals representing numerical data to provide a plurality of fuzzy signals representing fuzzified data; and

output neural means coupled to said input neural means for receiving and processing said plurality of fuzzy signals in accordance with a plurality of numerical fuzzy logic rule consequents in the form of singletons corresponding to unit-dimensional data to provide a plurality of numerical signals.

74. An artificial neural network as recited in claim 73, wherein said input neural means comprises a first plurality of artificial neurons.

75. An artificial neural network as recited in claim 73, wherein said output neural means comprises a second plurality of artificial neurons.

76. An artificial neural network as recited in claim 75, wherein said plurality of numerical fuzzy logic rule consequents comprise a plurality of neural weights associated with said second plurality of artificial neurons.

77. An artificial neural network as recited in claim 73, further comprising neural summing means coupled to said output neural means for receiving and summing said plurality of numerical signals.

78. An artificial neural network as recited in claim 77, wherein said neural summing means comprises an artificial neuron.

79. An artificial neural network for processing a plurality of input data in accordance with fuzzy logic, comprising:

a first plurality of artificial neurons which receive and fuzzify a plurality of input signals representing numerical data to provide a plurality of fuzzy signals representing fuzzified data; and

a second plurality of artificial neurons, coupled to said first plurality of artificial neurons, which receive and process said plurality of fuzzy signals in accordance with a plurality of numerical fuzzy logic rule consequents in the form of singletons corresponding to unit-dimensional data to provide a plurality of numerical signals.

80. An artificial neural network as recited in claim 79, wherein said plurality of numerical fuzzy logic rule consequents comprise a plurality of neural weights associated with said second plurality of artificial neurons.

81. An artificial neural network as recited in claim 79, further comprising an artificial neuron, coupled to said second plurality of artificial neurons, which receives and sums said plurality of numerical signals.

82. A computer-implemented method for processing a plurality of input data in accordance with fuzzy logic, comprising the computer-implemented steps of:

receiving a plurality of input signals representing numerical data;

fuzzifying said plurality of input signals to provide a plurality of fuzzy signals representing fuzzified data; and

processing said plurality of fuzzy signals in accordance with a plurality of numerical fuzzy logic rule consequents in the form of singletons corresponding to unit-dimensional data to provide a plurality of numerical signals.

83. A method as recited in claim 82, wherein said step of fuzzifying said plurality of input signals to provide a plurality of fuzzy signals representing fuzzified data comprises fuzzifying said plurality of input signals with a first plurality of artificial neurons.

84. A method as recited in claim 82, wherein said step of processing said plurality of fuzzy signals in accordance with a plurality of fuzzy logic rule consequents which are numerical to provide a plurality of numerical signals comprises processing said plurality of fuzzified data signals with a second plurality of artificial neurons.

85. A method as recited in claim 82, wherein said step of processing said plurality of fuzzy signals in accordance with a plurality of fuzzy logic rule consequents which are numerical to provide a plurality of numerical signals comprises processing said plurality of fuzzified data signals with a second plurality of artificial neurons and a plurality of neural weights associated therewith.

86. A method as recited in claim 82, further comprising the step of summing said plurality of numerical signals.

87. A method as recited in claim 86, wherein said step of summing said plurality of numerical signals comprises summing said plurality of numerical signals with an artificial neuron.