Open-Circuit Fault Diagnosis of Z-Source Inverter via Deep Neural Network

In this paper, issues of open-circuit fault diagnosis are solved for Z-source-inverter via the deep-sparse-autoencoder, which is one type of the various deep neural networks. Firstly, the preliminary of Z-source-inverter is introduced briefly. On the basis of the introduction of Z-source inverter, MATLAB/Simulink is utilized for the development of the open-circuit fault model of the Z-source inverter, and the output voltages of diverse fault cases are recorded for the training and test of the deep neural network based fault diagnosis. In the end, the superiority and effectiveness of the presented fault diagnosis approach can be verified by several comparisons with other methods on simulation experiments.