With the development of national economy , the demand for energy increase fast ,and wind energy become the fastest growing new energy for its clean ,efficient and renewable. As the rapid development of wind power generation, operation and maintenance of wind farms become one of the important issues that we will have to face. If a failure can be predicted before it comes to be true, we can make the maintenance plan and schedule maintenance personnel and supplies in time, so the loss caused by faults can be reduced. Therefore, it is very important to study on the fault prediction algorithm of key components in wind turbine.This paper introduces the research actuality of the domestic and international fault detection algorithm of wind turbine, along with the theory of support vector machine(SVM) and statistical process control (SPC)chart. We set up the fault prediction algorithm for blade based on improved SPC,and algorithm for wind turbine gearbox and generator based on least squares SVM.At last, this paper apply the fault prediction algorithms to the actual condition monitoring data in a wind farms, it proves the possibility and availability about the fault prediction algorithms. |