Claims
- 1. A fuzzy logic based model assessment system for assessing at least one model of physical phenomena using measurement residual values representing a difference between a measured data sequence corresponding to said physical phenomena and an expected data sequence corresponding to said model to be assessed, said system comprising:a feature identification module for identifying at least one feature present in said measurement residual values, and for generating at least one feature amplitude value and at least one feature amplitude standard deviation value; an anomaly characterization module for characterizing said at least one feature in at least one membership class based upon said at least one feature amplitude value, and for generating class membership interval values for each said at least one membership class based upon said at least one feature amplitude standard deviation value, and wherein said class membership interval values represent a range of degrees of membership in each said at least one membership class; and a hypothesis formulation and evaluation module for determining at least one mismodeling hypothesis by applying fuzzy inferencing to said class membership interval values, and for generating at least one hypothesis certainty value representing a degree of certainty of said at least one mismodeling hypothesis.
- 2. The system of claim 1 wherein said physical phenomena includes a moving contact, wherein said at least one model is a contact tracking model, and wherein said at least one feature includes at least one tracking anomaly.
- 3. The system of claim 2 wherein said at least one tracking anomaly includes a jump feature representing a discontinuity in a measured tracking signal and includes a drift feature representing a generally linear drift of said measured tracking signal.
- 4. The system of claim 1 wherein said at least one membership class includes one of null, weak, moderate and strong.
- 5. The system of claim 1 wherein said class membership interval values include an upper limit representing a greatest possible extent to which said at least one feature belongs to each said at least one membership class and includes a lower limit value representing a smallest necessary extent to which said at least one feature belongs to each said at least one membership class.
- 6. The system of claim 1 wherein said at least one hypothesis certainty value includes hypothesis certainty interval values representing a range of certainty for said at least one mismodeling hypothesis.
- 7. The system of claim 1 wherein said hypothesis formulation and evaluation module includes:a knowledge base including a plurality of rules for inferring at least one mismodeling hypotheses based upon features and membership classes; and an inferencing module for applying at least one rule from said knowledge base to said class membership interval values, for selecting at least one selected mismodeling hypothesis based upon said class membership interval values and said at least one rule, and for generating hypothesis certainty interval values representing a range of certainty for said at least one selected mismodeling hypothesis.
- 8. The system of claim 7 wherein said inferencing module applies a plurality of rules that result in the same said at least one mismodeling hypothesis with different said hypothesis certainty interval values, and further including an aggregation module for aggregating said hypothesis certainty interval values to generate a composite hypothesis certainty interval value for said at least one mismodeling hypothesis.
- 9. The system of claim 1 wherein said feature identification module, said anomaly characterization module, and said hypothesis formulation and evaluation module are implemented on a computer.
- 10. A fuzzy inference system for use with a model assessment system to infer possible causes of mismodeling in at least one model of physical phenomena based upon at least one feature identified in measurement residual values, wherein said at least one feature is represented by at least one feature amplitude value and at least one feature amplitude standard deviation value, said system comprising:an anomaly characterization module for characterizing said at least one feature in at least one membership class based upon said at least one feature amplitude value, and for generating class membership interval values for each said at least one membership class based upon said at least one feature amplitude standard deviation value, and wherein said class membership interval values represent a range of degrees of membership in each said at least one membership class and include an upper limit representing a greatest possible extent to which said at least one feature belongs to each said at least one membership class and includes a lower limit value representing a smallest necessary extent to which said at least one feature belongs to each said at least one membership class; and a hypothesis formulation and evaluation module for determining at least one mismodeling hypothesis by applying fuzzy inferencing to said class membership interval values, and for generating at least one hypothesis certainty value representing a degree of certainty of said at least one mismodeling hypothesis.
- 11. The system of claim 10 wherein said physical phenomena includes a moving contact, wherein said at least one model is a contact tracking model, and wherein said at least one feature includes at least one tracking anomaly.
- 12. The system of claim 11 wherein said at least one tracking anomaly includes a jump feature representing a discontinuity in a measured tracking signal and includes drift feature representing a generally linear drift of said measured tracking signal.
- 13. The system of claim 10 wherein said at least one membership class includes null, weak, moderate or strong.
- 14. The system of claim 10 wherein said at least one hypothesis certainty value includes hypothesis certainty interval values representing a range of certainty for said at least one mismodeling hypothesis.
- 15. The system of claim 10 wherein said hypothesis formulation and evaluation module includes:a knowledge base including a plurality of rules for inferring mismodeling hypotheses based upon features and membership classes; and an inferencing module for applying at least one rule from said knowledge base to said class membership interval values, for selecting at least one selected mismodeling hypothesis based upon said class membership interval values and said at least one rule, and for generating hypothesis certainty interval values representing a range of certainty for said at least one selected mismodeling hypothesis.
- 16. The system of claim 10 wherein said anomaly characterization module and said hypothesis formulation and evaluation module are implemented on a computer.
- 17. A method of assessing at least one model of physical phenomena in response to a measured data sequence representing a signal caused by said physical phenomena in the presence of noise, said method comprising:providing measurement residual values representing a difference between said measured data sequence and an expected data sequence corresponding to said model to be assessed; identifying at least one feature present in said measurement residual values; generating at least one feature amplitude value and at least one feature amplitude standard deviation value for said at least one feature identified; characterizing said at least one feature in at least one membership class based upon said at least one feature amplitude value; generating class membership interval values for each said at least one membership class based upon said at least one feature amplitude standard deviation value, and wherein said class membership interval values represent a range of degrees of membership in each said at least one membership class; determining at least one mismodeling hypothesis by applying fuzzy inferencing to said class membership interval values, wherein said mismodeling hypothesis represents a possible cause of mismodeling; and generating at least one hypothesis certainty value representing a degree of certainty of said at least one mismodeling hypothesis.
- 18. The method of claim 17 wherein said physical phenomena includes a moving contact, wherein said at least one model is a contact tracking model, and wherein said at least one feature includes at least one tracking anomaly.
- 19. The method of claim 18 wherein said at least one tracking anomaly includes a jump feature representing a discontinuity in a measured tracking signal and includes drift feature representing a generally linear drift of said measured tracking signal.
- 20. The method of claim 17 wherein said at least one membership class includes null, weak, moderate or strong.
- 21. The method of claim 17 wherein said class membership interval values include an upper limit representing a greatest possible extent to which said at least one feature belongs to each said at least one membership class and includes a lower limit value representing a smallest necessary extent to which said at least one feature belongs to each said at least one membership class.
- 22. The method of claim 17 wherein the step of generating said at least one hypothesis certainty value includes generating hypothesis certainty interval values representing a range of certainty for said at least one mismodeling hypothesis.
- 23. The method of claim 17 wherein the steps of determining at least one mismodeling hypothesis and generating said hypothesis certainty interval values include:providing a knowledge base including a plurality of rules for inferring mismodeling hypotheses based upon features and membership classes; applying at least one rule from said knowledge base to said class membership interval values; selecting at least one selected mismodeling hypothesis based upon said class membership interval values and said at least one rule; and generating hypothesis certainty interval values representing a range of certainty for said at least one selected mismodeling hypothesis.
- 24. The method of claim 23 wherein said inferencing module applies a plurality of rules that result in the same said at least one mismodeling hypothesis with different said hypothesis certainty interval values, and further including the step of:aggregating said hypothesis certainty interval values to generate a composite hypothesis certainty interval value for said at least one mismodeling hypothesis.
- 25. A fuzzy inference system for use with a model assessment system to infer possible causes of mismodeling in at least one model of physical phenomena based upon at least one feature identified in measurement residual values, wherein said at least one feature is represented by at least one feature amplitude value and at least one feature amplitude standard deviation value, said system comprising:an anomaly characterization module for characterizing said at least one feature in at least one membership class based upon said at least one feature amplitude value, and for generating class membership interval values for each said at least one membership class based upon said at least one feature amplitude standard deviation value, and wherein said class membership interval values represent a range of degrees of membership in each said at least one membership class and include an upper limit representing a greatest possible extent to which said at least one feature belongs to each said at least one membership class and includes a lower limit value representing a smallest necessary extent to which said at least one feature belongs to each said at least one membership class; and a hypothesis formulation and evaluation module for determining at least one mismodeling hypothesis by applying fuzzy inferencing to said class membership interval values, and for generating at least one hypothesis certainty value representing a degree of certainty of said at least one mismodeling hypothesis, wherein said hypothesis formulation and evaluation module includes (i) a knowledge base including a plurality of rules for inferring mismodeling hypotheses based upon features and membership classes; (ii) an inferencing module for applying at least one rule from said knowledge base to said class membership interval values, for selecting at least one selected mismodeling hypothesis based upon said class membership interval values and said at least one rule, and for generating hypothesis certainty interval values representing a range of certainty for said at least one selected mismodeling hypothesis, wherein said inferencing module applies a plurality of rules that result in the same said at least one mismodeling hypothesis with different said hypothesis certainty interval values; and (iii)an aggregation module for aggregating said hypothesis certainty interval values to generate a composite hypothesis certainty interval value for said at least one mismodeling hypothesis.
- 26. A method of assessing at least one model of physical phenomena in response to a measured data sequence representing a signal caused by said physical phenomena in the presence of noise, said method comprising:providing measurement residual values representing a difference between said measured data sequence and an expected data sequence corresponding to said model to be assessed; identifying at least one feature present in said measurement residual values; generating at least one feature amplitude value and at least one feature amplitude standard deviation value for said at least one feature identified; characterizing said at least one feature in at least one membership class based upon said at least one feature amplitude value; generating class membership interval values for each said at least one membership class based upon said at least one feature amplitude standard deviation value, and wherein said class membership interval values represent a range of degrees of membership in each said at least one membership class; determining at least one mismodeling hypothesis by applying fuzzy inferencing to said class membership interval values, wherein said mismodeling hypothesis represents a possible cause of mismodeling; generating at least one hypothesis certainty value representing a degree of certainty of said at least one mismodeling hypothesis; and using said at least one mismodeling hypothesis and said at least one hypothesis certainty value to assess said at least one model and generate a tactical picture.
STATEMENT OF GOVERNMENT INTEREST
The invention described herein may be manufactured and used by or for the Government of the United States of America for governmental purposes without the payment of any royalties thereon or therefore.
US Referenced Citations (12)