Algorithm of Data Mining and its Application in Fault Diagnosis for Wind Turbine

Because more and more characteristic parameters are used to describe machine vibration, few and more parameters selected will usually affect the right ration of diagnosis. In this paper an algorithm of data mining based on fuzzy clustering is approached, which is used to mine the optimal characteristic parameters from the parameters describing the vibration wave of gear cases of wind turbines. The detailed steps of data mining method were introduced. The characteristic parameters mined were used as optimal feature eigenvectors for diagnosing faults of gear cases. Applied example was shown that the result of fault diagnosis has been proved to be reliable and accurate.