Development of diagnosis system for HV motor insulation based on multi-parameter fusion

On the base of studying the relationship between the working condition of HV motor and parameters of partial discharge (PD), a diagnosis system for motor insulation based on multisource information fusion and multi-layer logical reasoning was developed. The system mainly relies on the data from PD. Other data from temperature and vibration sensors is also used to assist the analysis. The diagnostic method, which is from the part to the whole, from analysis to synthesis is adopted in the system. Artificial neural network and expert system are used, the system achieves fault diagnosis and condition assessment of HV motor insulation with on-line monitoring system of HV motor insulation together.