Research On Fuzzy Diagnostic Techniques For Gearbox’s Faults Of Megawatt Wind Generator

With rapid development of wind turbine industry, the function of more stability andhigh efficiency are required in wind turbine generator system. In order to decrease thefailure rate of gearboxes, effective methods need to be searched. The vibration signals ofwind turbine gearbox have some features which include nonstationarity, diversificationand so on. Early fault diagnosis methods should be searched deeply based on windturbine gearbox’s failure mechanism and fault signal characteristics. The failure modesand typical failure modes of Wind turbine gearboxes are summarized by simulation. Thetime domain diagnostics and the frequency domain diagnostics which based on signalprocessing are introduced. Fundamental of the diagnostics are studied. Specific Scope oftime and frequency domain factors using in some methods be mentioned is analyzed.Based on the research of wind turbines gearbox failure and vibration mechanism, windturbines gearbox vibration signal of four different conditions by Germany advancedvibration signal acquisition and analysis equipment and frequency domain analysis of thevibration signal can be used to process the signal. Based on twelve time domain andfrequency domain characteristic parameters which can reflect wind turbine gearboxfailure, we can conduct a multi-parameter identification and establish a typical datasample of wind turbine gearbox failures. Base the typical data sample of wind turbinegearbox failures, the logic and the algorithm of fuzzy clustering analysis algorithm ineach step can be studied, and the optimum algorithm could be selected for programming.Finally, the fuzzy clustering analysis method is applied to fault diagnosis of wind turbinegearboxes. Based on the fuzzy clustering analysis results, it is proved that the proposedfault diagnosis method is feasible and can effectively be analysed and diagnosed gearboxfault.