Fuzzy Information Fusion Algorithm of Fault Diagnosis Based on Similarity Measure of Evidence

In this paper, a fuzzy information fusion method of fault diagnosis based on evidence similarity measure is presented. First, because of the fuzziness of information received by sensors, membership functions are introduced to describe the fault template mode in model database and features extracted from sensor observations; then the degrees of matching between them are obtained using random set model of fuzzy information, which can be transformed into BPAs. Second, cosine similarity measure of evidence is introduced to compute confidence degree of evidence. Finally, original evidences are modified according to the confidence degree. The diagnosis results of rotor system show that the proposed method can improve the accuracy of decision-making.