Incomplete Overhaul Strategy Study of Open-Circuit Faults in Doubly-Fed Wind Turbine Converters
To solve the problem that open-circuit faults are prone to reoccur during fault overhaul,an incomplete overhaul strategy of open-circuit faults in doubly-fed wind turbine converters based on the improved maximum entropy algorithm is proposed.The infrared thermal image of the state of doubly-fed wind turbine converter is collected by using infrared thermal imager,and the recognition of the image details is improved by filtering and enhancement processing.Innovatively,the optimal segmentation threshold is determined by improving the maximum entropy algorithm,the target and background of the image are segmented,and the target part is selected for fault overhaul.Convolutional neural network is used to extract image features and dimensionality reduction.A classifier is used to estimate the likehood of each fault category to realize incomplete overhaul of open-circult faults in doubly-fed wind turbine converters.The study results show that the proposed strategy can accurately overhaul the open-circuit faults of doubly-fed wind turbine converters with overhaul accuracy,precision,recall and F1 values of 0.8,1.0,1.0 and 0.86,respectively.Compared with the existing strategy,the proposed strategy has high overhaul accuracy and reliability,and it can solve the problem of the reoccurrence of open-circuit faults during fault overhaul.
万方数据
维普
