Wavelet transform based open circuit fault diagnosis in the converter used in wind energy systems

This paper is an attempt to develop a novel method of training the ANN for fault diagnosis technique of the open circuit faults in the PMSG wind turbine connected converters. The training portion of any artificial intelligence methods is time taking and needs lots of tuning for creating the black box. Thus the training portion of the fault diagnosis in the ANN structure has to be optimized. The Wavelet transform from the signal of interest, which is the response of the fault, occurred in the input voltage is taken into consideration as the feature to be extracted to be given as the input to the ANN for training. And the output of the ANN, which has to be trained, is taken as the fault status, which corresponds to the Wavelet coefficients. Thus the time taken for the training would get reduced and the accuracy of the prediction would get improved while the Wavelet transform based method is used. The open circuit faults are found for each switch in the converter. Matlab/Simulink and Wavelet Toolbox
^™
based implementation is carried out and the results for training and testing the diagnosis system is tabulated and inferred.