Fault diagnosis method of the locomotive brake based on wavelet analysis

Proposed a locomotive brake fault Diagnostic expert system approach, by using wavelet multiscale edge detection principle to get fault symptom extraction, and combined with the knowledge representation method of more signs of weighted fuzzy production rule, in order to make the failure symptoms and findings fuzzy and weight conclusions of fuzzy fault symptoms, weight based, fuzzy rule inference algorithm by further diagnostic reasoning, fault diagnosis of locomotive brake to solve the uncertainty of the process. A practical example indicate that the diagnosis system can improve the real-time and veracity of the diagnosis for air brake of diesel locomotive effectively.