The adaptive fuzzy - support vector machine for fault detection and isolation in wind turbine
This paper concerns the problem of fault diagnosis in wind turbines. Motived by SupportVector Machines (SVM) method and Fuzzy Logic algorithm, a novel procedure is derived to provide for the wind turbine diagnosis. Since, the conventional SVM classifier with fixed parameters cannot bring performance of high accuracy and fast reflex. In this work, the proposed FDI strategy has raised the problem of congeal the parameters of classifiers after learning, this novel strategy based on the evaluating of error of classification to adjust the parameters of classifier w and b in real time using the fuzzy logic. This principal allow for achieve a new classifier online, which is able to process the new data comes from measuring sensors. The Different parts of the process were investigated, including actuators, sensors and process faults. With duplicated sensors, have detected sensor faults in blade pitch positions, generator and rotor speeds rapidly, but under specific constraints on the fault. all Process faults mainly concerned friction in the wind turbine, which might cause it damage. The fault could be detected under constraints of high magnitude error. the comparing Our results with the conventional SVM classifier indicate the value of our method
Keywords: Fault Detection and Isolation (FDI), wind turbine, classification, SupportVector Machine (SVM), Fuzzy logic
Y. Fadili, I. Boumhidi, "The adaptive fuzzy - support vector machine for fault detection and isolation
in wind turbine", Engineering Intelligent Systems, vol. 26 no. 1, pp. 35-43, 2018.