Machinery fault diagnosis method of HV circuit breaker based on EEMD and RBF neural network

HV circuit break is a kind of important switching equipment in the field of power system. In order to improve safety and reliability of power system, studies about fault diagnosis of high-voltage circuit breaker are needed, especially for mechanical fault. In the study field of mechanical fault diagnosis of HV circuit breaker, the diagnosis process includes three steps: signal acquisition, feature extraction and fault identification. The methods of Fault identification mainly can be divided as three aspects, it is model identification, signal identification and knowledge identification. In this article, the ensemble empirical mode decomposition (EEMD) is used for feature extraction, then the EEMD-characteristic entropy can be obtained. However, the frequency of the mechanical action of high-voltage circuit breaker is very few, the experimental data about EEMD-characteristic entropy is precious and highly depends on the existing samples. For classification issues, in this study, radial basis function (RBF) neural network which act as a recognition tool were used to fault diagnosis. The whole process of this research included: signal acquisition, feature extraction, fault diagnosis.