Fault diagnosis of VSG inverter under grid connection

Abstract

Under different working conditions, the output variable waveform corresponding to the same fault will be different. This article introduces the VSG inverter model under grid connection and verifies the correctness of the model. The three-phase current of the inverter bridge arm is selected as the research object, and the time-domain two-dimensional graph is formed as the And classify single IGBT and 2 IGBT open circuit faults, and encode various faults. eature vector; the deep convolutional neural network is established and the fault diagnosis is performed, and the feature vector is used as the input and the corresponding fault code is the output. The training simulation realizes and verifies the feasibility and accuracy of the fault diagnosis of the VSG inverter using the deep convolutional neural network.