Mechanical Fault Diagnosis of High Voltage Circuit Breakers Based on Phase Space Reconstruction and Improved GSA-SVM

The vibration signals generated by transmission and impact of circuit breaker mechanical components have chaotic characteristics, which are difficult to be analyzed with conventional signal processing methods. Firstly, the vibration signals are reconstructed into a high-dimensional space by mutual information method and Cao algorithm, and the permutation entropy is calculated as the feature vector. And then the support vector machine (SVM) is used to identify the mechanical fault types of circuit breakers. Finally, the PSO improved GSA hybrid algorithm is used to optimize the parameters of SVM, and the measured vibration signals of the circuit breakers are used to verify the results. The results show that the characteristics of circuit breaker vibration signals can be accurately extracted with combination of phase space reconstruction and permutation entropy. The PSO-GSA-SVM can quickly and effectively identify the fault types of circuit breakers, thus providing an effective solution to such problems as path distortion, energy leakage and mode overlap of existing diagnosis methods.