Fault Diagnosis And Condition Monitoring Of Cascaded Inverter

At present, the cascaded inverter has been widely used in drive fans, pumps and all kinds of large-scale machines. It has become one of the main means of saving the current electricity, improving the level of the production process automation and promoting technological progress. Compared with the traditional two levels inverter, the cascaded inverter has more power devices in the main circuit, so the possibility of the fault problem has greatly increased. The power component open-circuit fault of the cascaded inverter has little impact on the output voltage and current, so it is difficult to find. If the fault can’t be diagnosed in time, a lot of problems will be formed. However,the research aiming at the fault diagnosis of the cascaded inverter is few at present.Therefore, it is imperative to find a cascaded inverter fault diagnosis method with both efficiency and accuracy.The cascade inverter power device open fault is taken as the object of study in this paper. Firstly the structure and the working principle of the cascade inverter is analyzed,and the simulation model of cascade converters is constructed in MATLAB/Simulink module, Then the open circuit fault of the drive power devices is simulated respectively in the DC filter capacitors without considering the impact of aging factors and considering the impact of aging factor. The source of the failure characteristics of the signal is determined by analyzing the output signals in different situations of the cascade inverter. And then the wavelet analysis is used to extract its fault feature to compose the feature vector input of the neural networks.Secondly, Aiming at the shortcomings of the cascade inverter fault diagnosis by the loosing wavelet neural network, such as the slow network convergence and easy to fall into the local minimum value, the PSO-NN is proposed to diagnose the faults to improve the accuracy of the cascade inverter. At the same time, in order to avoid the premature convergence and the low iterative precision in the late time of the basic PSO algorithm, the speed weights and the acceleration coefficient are adjusted. The simulation results show that the improved algorithm has higher diagnostic accuracy and faster convergence speed than the traditional BP algorithm.Finally, the LabVIEW is used to design a cascade inverter fault diagnosis systembase on the PSO-NN fault diagnosis method. The experimental results show that the fault diagnosis system can diagnosis the faults of the cascaded converters effectively.