Research on Capacitor Aging Fault Diagnosis Method of Inverter

In order to solve the problems of multiple classification and subclassification, a feature extraction strategy combining adaptive white noise complete set empirical Mode decomposition and wavelet packet energy entropy was proposed. The parameters of least squares support vector machine are optimized by the improved Sparrow search algorithm. Firstly, the phase voltage signal was processed by CEEMDAN to obtain the modal component, and IMF was screened jointly according to the correlation coefficient and variance contribution rate. The IMF with noise was denoised and reconstructed, and the IMF without noise was formed into the pure IMF group. Then, WPEE with obvious fault characteristics was decomposed by wavelet packet analysis. Secondly, LSSVM parameters are optimized by using Iterative chaotic mapping and the improved SSA based on random walk strategy, and the diagnostic model is established. Finally, Z-source inverter is taken as an example for verification. The results show that the proposed method can extract capacitor aging fault features quickly and effectively, and the diagnosis method is faster and the fault recognition rate is higher.