Electromechanical actuator interpretable fault diagnosis method considering language fuzzy information

Abstract

This invention relates to the field of electromechanical equipment health management technology, specifically to an interpretable fault diagnosis method for electromechanical actuators that considers linguistic fuzzy information. The method includes: determining a set of fault modes and a set of fault features; based on expert knowledge, providing the probability distribution of each fault feature at different reference levels in the form of interval gray numbers, forming an interval gray number table; calculating the total uncertainty of each interval gray number and converting it into confidence in evidence reasoning, while simultaneously converting the interval gray numbers into basic probability assignments for each fault mode, generating an evidence set containing confidence and basic probability assignments; constructing an initial model of a confidence rule base using the evidence set; optimizing the parameters of the initial model using an adaptive evolutionary strategy algorithm with interpretable constraints on the projection covariance matrix, obtaining an optimized model; and based on the fault feature data of the electromechanical actuator to be diagnosed, obtaining the confidence distribution of each fault mode through the optimized model, thereby achieving fault identification and diagnosis.

Images