基于径向基函数分类器的断路器机械故障诊断

The mechanical performance damage is the dominant factor causing the fault of circuit breaker; therefore, the study on mechanical fault diagnosis for circuit breaker has very important theoretical value and practical significance. For achieving effective diagnosis for the mechanical fault of circuit breaker, a kind of identification and classification algorithm based on Gaussian radial basis function is proposed in this paper. This algorithm takes the normalized energy spectrum in the whole effective band of signal as the feature vector, in which every frequency component corresponds to one characteristic dimension without involving time frequency domain segmentation. Therefore, it has stronger adaptability without considering segmentation foundation and rationality. In addition, the constructed fault diagnosis system has a parameter correction and optimization function. Finally, the effectiveness of the mechanical fault diagnosis algorithm of the circuit breaker proposed in this paper is verified by the study of experiment.