电子机械论文电力电子系统故障诊断方法浅析（Analysis of the fault diagnosis method of the electronic mechanical power electronic system）.docVIP

电子机械论文电力电子系统故障诊断方法浅析（Analysis of the fault diagnosis method of the electronic mechanical power electronic system） The key of fault diagnosis is to extract fault characteristics. Fault characteristic is the signal that reflects the signal of fault symptom, and reflects the type, position and degree of equipment and system. The difference method of the fault diagnosis method according to feature extraction, can be divided into spectral analysis method, based on the dynamic mathematical model of the system, using the method of pattern recognition method based on neural network method, expert system method, wavelet transform method and genetic algorithm etc.. These methods will be described in detail below. Spectrum analysis method in fault diagnosis In fault diagnosis, the signal processing method is spectrum analysis. Commonly used Fourier spectrum, Walsh spectrum, in addition to filtering, correlation analysis. The purpose of spectrum analysis is that the signal contains noise, in order to extract features, the time domain waveform of the fault signal can not clearly reflect the characteristics of the fault. The key point in power electronic circuits containing fault information signal usually has a periodic, so it can be the fault waveform transform in the time domain to the frequency domain by Fourier transform, to highlight the fault features, fault diagnosis. Fu Liye transform is to decompose a periodic function into sinusoidal components of various frequencies. Similarly, the Walsh transform is a decomposition of a function into a set of Walsh functions. Adaptive filtering is a statistical method for digital signal processing, and it does not need prior knowledge of statistics signal of order one or two, directly using the observational data, through some parameters can change the filter, and the output of the adaptive filter can change the automatic tracking signal characteristics. In the fault diagnosis of power electronic system, adaptive processing can be us