A Wind Turbine Fault Diagnosis Method Using Adaptive Weighting Algorithm and D-S Evidence Theory

Since the system of Wind Turbine is complex and the relationships between faults and symptoms are not simple or linear, ordinary detection can not meet the diagnostic requirements, Data Fusion is an important problem in the context of wireless sensor networks. This paper proposed a data fusion method on Wind Turbine's state monitoring and fault diagnosis. First, take Adaptive Weighting Algorithm into signal data fusion which reduces data redundancy and energy consumption of network; then, take Gauss membership function to obtain the Basic Probability Assignment which can improve the data reliability in D-S Evidence Theory. The improved D-S Evidence Theory in feature data fusion can enhance the ability of fault identification. The experimentation shows the proposed method has a high diagnostic accuracy and reliability.