Fault diagnosis based on fuzzy information multi-level fusion

To enhance the reliability of synthetical fault diagnosis, a fault diagnosis method based on fuzzy information multi-level fusion was developed. Synthetical fault diagnosis was carried out after fusing again the diagnostic data in each level based on conventional fault diagnosis method. Tests on ship main power system show that the method is more accurate and effective comparing with conventional method. Even in the case of failures of the local detection sensor, the method can still make correct fault diagnosis, which gives a beneficial reference to enhance the reliability of fault diagnosis for large systems.