Applications of Fuzzy Multilayer Support Vector Machines in Fault Diagnosis and Forecast of Electric Power Equipment

In order to improve the utilization rate of power equipment and service life, reduce the economic loss of the power grid and overcome the shortcomings of traditional threshold determination methods, this paper proposed a power equipment fault diagnosis method based on fuzzy multilayer support vector machines. The new method mines equipment feature information by the fuzzy logic algorithm, trains many different support vector machines using multilayer feature information as input and determines devices' fault class by the majority voting method. Through the application of the algorithm in fault diagnosis of circuit breaker and transformer, it's obvious that the new method greatly overcomes abnormal data's influence caused by the noise and the abnormal operation of sensors and improves the fault diagnosis accuracy of equipment.