Fault prediction algorithm for offshore wind energy conversion system based on machine learning

In view of the low accuracy and long time sum difference of fault prediction methods in offshore wind power conversion system, a fault prediction algorithm for offshore wind power conversion system is designed based on SVM method in machine learning. The fault data of offshore wind power conversion system are preprocessed by principal component analysis, and the fault characteristics of offshore wind power conversion system are extracted by leftflush method. Based on the principle of SVM, the fault prediction LSSVM model of offshore wind power conversion system is constructed. The test results show that this method has good fault prediction effect for offshore wind power conversion system, and can effectively improve the accuracy of fault prediction and shorten the time of fault prediction.