Fuzzy logic estimator for minimizing signal measurement inaccuracy effects in a signal processing system

• US 6,662,140 B2
• Filed: 06/12/2001
• Issued: 12/09/2003
• Est. Priority Date: 06/12/2001
• Status: Active Grant

• Current Assignee: Rolls-Royce Canada Ltd (Rolls-Royce Holdings Plc)
• Original Assignee: Rolls-Royce Canada Ltd (Rolls-Royce Holdings Plc)
• Inventors: Martis, Dan
• Primary Examiner(s): Barlow, John
• Assistant Examiner(s): Cherry, Stephen J.
• Application Number:
  US09/879,251
• Publication Number:
  US 20020188429A1
• Time in Patent Office: 910 Days
• Field of Search: 700/50, 700/52, 706/1, 706/2, 706/4, 706/8, 706/22, 702/179, 702/181, 702/189, 702/190, 702/194
• US Class Current: 702/179
• CPC Class Codes: G01D 3/02 with provision for altering or correcting the law of variation
• CPC Class Codes: G05B 13/0275 using fuzzy logic only

Abstract

A fuzzy logic estimator for minimizing signal measuring inaccuracy effects in a signal processing system preferably includes a microprocessor-based system operable to receive a number of measured signal values and estimate a solution to an overdetermined system of equations that minimizes differences between the measured signal values and corresponding model values. The fuzzy logic solution estimate process includes assigning a probability distribution to delta values representing differences between the measured signal values and corresponding model values to form a corresponding number of probability distribution functions, associating at least some of the probability distribution functions with each equation of a system of equations defining a number of unknown parameter values, solving the system of equations for a domain of possible solutions, and determining a unique solution for the unknown parameter values from the domain of possible solutions.

Claims

1. A method of minimizing signal measurement inaccuracy effects in a signal processing system, comprising:

assigning a probability distribution to each of a first number of delta values to form a corresponding first number of probability distribution functions, said delta values representing differences between pairs of measured signal values and corresponding model values;

associating at least some of said first number of probability distribution functions with each equation in a system of equations defining a second number of unknown parameter values;

solving said system of equations for a domain of possible solutions; and

determining a unique solution for said second number of unknown parameter values from said domain of possible solutions.

2. The method of claim 1 wherein said probability distribution is a normal probability distribution.

3. The method of claim 1 wherein said probability distribution is a trapezoidal probability distribution.

4. The method of claim 1 wherein said probability distribution is a triangular probability distribution.

5. The method of claim 1 wherein said probability distribution is a rectangular probability distribution.

6. The method of claim 1 wherein said probability distribution is a nonsymmetrical probability distribution.

7. The method of claim 1 wherein the solving step includes forming said domain of possible solutions as a product of probability distribution functions associated with each equation in said system of equations.

8. The method of claim 1 wherein the determining step includes determining said unique solution as a function of a maximum value of said domain of possible solutions.

9. The method of claim 1 wherein the determining step includes determining said unique solution as a centroid of said domain of possible solutions.

10. The method of claim 1 wherein said first number is greater than said second number such that said system of equations represents an over-determined system of equations.

11. The method of claim 1 wherein the solving step includes defining said domain of possible solutions as an intersection of said probability distribution functions associated with each equation in said system of equations.

12. A method of minimizing signal measurement inaccuracy effects in a signal processing system, comprising:

measuring a plurality of signal values;

computing a first number of delta values each representing a difference between one of said plurality of signal values and a corresponding model value;

assigning a probability distribution to each of said first number of said delta values to form a corresponding first number of probability distribution functions;

associating at least some of said first number of probability distribution functions with each equation in a system of equations defining a second number of unknown parameter values;

solving said system of equations for a domain of possible solutions; and

determining a unique solution for said second number of unknown parameter values from said domain of possible solutions.

13. The method of claim 12 wherein the determining step includes determining said unique solution as a function of a maximum value of said domain of possible solutions.

14. The method of claim 12 wherein the determining step includes determining said unique solution as a centroid of said domain of possible solutions.

15. The method of claim 12 wherein said first number is greater than said second number such that said system of equations represents an over-determined system of equations.

16. The method of claim 12 wherein said probability distribution is one of a normal probability distribution, a trapezoidal probability distribution, a triangular probability distribution and a rectangular probability distribution function.

17. The method of claim 12 wherein said probability distribution is a non-symmetrical probability distribution.

18. The method of claim 12 wherein the solving step includes defining said domain of possible solutions as an intersection of said probability distribution functions associated with each equation of said system of equations.

19. A system for minimizing signal measurement inaccuracy effects in a signal processing system, comprising:

a plurality of sensors producing a corresponding plurality of signal values indicative of operating conditions of a physical system; and

a signal processing system receiving said plurality of signal values, said signal processing system including;

means for assigning a probability distribution to each of a first number of delta values to form a corresponding first number of probability distribution functions, said delta values representing differences between pairs of measured signal values and corresponding model values;

means for associating at least some of said first number of probability distribution functions with each equation of a system of equations defining a second number of unknown parameter values;

means for solving said system of equations for a domain of possible solutions; and

means for determining a unique solution for said second number of unknown parameter values from said domain of possible solutions.

20. The system of claim 19 wherein said first number is greater than said second number such that said system of equations defines an over-determined system of equations.

21. The system of claim 19 wherein said probability distribution is one of a normal probability distribution, a trapezoidal probability distribution, a triangular probability distribution and a rectangular probability distribution function.

22. The system of claim 19 wherein said probability distribution is a non-symmetrical probability distribution.

23. A system for minimizing signal measurement inaccuracy effects in a signal processing system, comprising:

a first circuit receiving a plurality of measured signal values and producing a first number of delta values each as a difference between one of said plurality of measured signal values and a corresponding model value; and

a second circuit assigning a probability distribution to each of said first number of delta values to form a corresponding first number of probability distribution functions, said second circuit associating at least some of said first number of probability functions with each equation of a system of equations defining a second number of unknown parameter values and solving said system of equations for a domain of possible solutions, said second circuit producing a unique solution for said second number of unknown parameters based on said domain of possible solutions.

24. The system of claim 23 wherein said second circuit is configured to solve said system of equations by defining said domain of possible solutions as an intersection of said probability distribution functions associated with each equation of said system of equations.

25. The system of claim 23 further including a third circuit operable to predict model parameters based on said second number of unknown parameters defining said unique solution.

26. The system of claim 25 further including a memory storing said predicted model parameters.

27. The system of claim 23 wherein said second circuit is configured to determine said unique solution as a function of a maximum of said domain of possible solutions.

28. The system of claim 23 wherein said second circuit is configured to determine said unique solution as a function of a centroid of said domain of possible solutions.

29. The system of claim 23 wherein said first number is greater than said second number such that said system of equations defines an over-determined system of equations.

30. The system of claim 23 wherein said probability distribution is one of a normal probability distribution, a trapezoidal probability distribution, a triangular probability distribution and a rectangular probability distribution function.

31. The system of claim 23 wherein said probability distribution is a nonsymmetrical probability distribution.