A fault diagnosis method based on fuzzy preference relation and D-S evidence theory

Abstract

The invention provides a method based on fuzzy preference relation and D-S evidence theory. The first step is to obtain the basic probability distribution function of multi-sensor information in mechanical equipment monitoring; In the second step, the credibility of the evidence is obtained by the cosine angle function, that is, the credibility matrix. In the third step, we construct the uncertainty matrix based on Deng entropy variance, measure the uncertainty of evidence, and obtain the preference relationship between the evidence. In the fourth step, the fuzzy preference relation matrix isconstructed by using the preference relation obtained in the previous step, and the ranking value instead of BPA is generated. In the fifth step, the weighted average evidence is obtained by modifyingthe weight vector through the confidence matrix and the ranking value. In the step 6 Fusion of weighted average evidences by Dempster combination rule for n-1 times, decisions are made once the results of the fusion are obtained. The present invention proposes a novel conflict evidence correction technique for obtaining modified weighted average evidence.

Description

technical field [0001] The invention belongs to the field of fault monitoring and diagnosis of large electromechanical equipment systems, and in particular relates to a fault intelligent decision-making method based on fuzzy preference relationship and D-S evidence theory. Background technique [0002] With the development of science and technology and the progress of society, various types of large electromechanical equipment have been widely used in engineering. Once these large-scale equipment fails, it may cause property loss, and even cause casualties in serious cases. For the research of equipment fault diagnosis methods, the complexity of the electromechanical system structure and the diversity of the working environment lead to more and more types of fault-related information, coupled with the limitations of sensor measurement accuracy, imperfect experimental methods and fault occurrence Factors such as the randomness of the location make the fault signals obtained ...

AI Technical Summary

Problems solved by technology

Lack of robustness is considered as a limitation of D-S evidence theory, which will greatly limit its application in decision-making systems for mechanical fault diagnosis

Images