A New Method of Obtaining BPA and Application to the Bearing Fault Diagnosis of Wind Turbine

Because of high computational complexity of neural network to obtain the basic probability assignment of evidence theory, this paper proposed a method of rough set theory based on random set and BP neural network to obtain the basic probability assignment. In the framework of random set, this paper makes use of the ability of rough set’s attribute reduction to reduce the neural network input dimension. Thus, this method is applied to bearing fault diagnosis of wind turbine, and better results are to be achieved.