The Fault Diagnosis of Wind Turbine Bearing Based On ISAPSO-SVM

摘要

Bearing is an important part of wind turbine gearbox. Its state will directly affect the safe and stable operation of wind turbine, so it's very important to diagnose the fault of bearing. The support vector machine (SVM) can do the classification work, but its classification accuracy is not good enough. Basic particle swarm optimization (PSO) has disadvantages about premature convergence and particles are easy to fall into local optimal solution. This paper proposed an effective fault diagnosis algorithm called ISAPSO-SVM (using improved simulated annealing particle swarm optimization to optimize SVM). Experimental results showed that, based on applying ISAPSO to solve the parameter selection problems in SVM, UCI database were used to do a series of classification experiments and the classification accuracy of ISAPSO-SVM was verified. Finally, the ISAPSO-SVM model was applied in wind turbine bearing diagnosis and a good result was achieved.