Research And Development Of An Fault Diagnosis System Of Wind Turbine Gearbox Based On Embeded Platform

With the severe lack of traditional energy, wind power as a kind of emerging renewable energy is getting more attention. Compared with other green energy, wind energy has advantages of less limitation of resources and good industry bases. With the impact of Mexico Gulf Oil Spill and Japan's Nuclear Crisis, wind energy will get greater development. Accompany the rapid development of wind installed capacity, single wind turbine capacity and cost are increasing in the meantime. How to monitor and maintain such a big expensive machine is an equal important issue.Wind generators usually work in very harsh environment conditions, which often leads to failure, the longest downtime is caused by gearbox, out of the consideration of security and high efficiency of wind generators, developing a set of online monitoring and fault diagnosis system for the gearbox of wind generators seems to be very important. This paper develops such a system based on laboratory conditions.The specialization of wind generator determines the specialization of the condition monitoring and diagnosis system. First of all, the wind turbine is far from ground and the cabin space is limited, so the monitoring equipment should not be too big or bulky, we adopt sbRIO which is an embedded veneer computer as the monitoring platform, it is compact flexible and very convenient to be moved and installed. Secondly, due to the time varying speed of wind, common fault diagnosis analysis method cannot meet the requirement of the unsteady signal analysis, this paper study that how order analysis can reduce the speed fluctuation impact, and through experiments we verify it is more suitable for unsteady wind generator gearbox signal analysis than envelop power analysis.Support vector machine is more suitable for fault diagnosis system of wind turbine for it needs fewer samples and has good generalization ability. This paper researches how to build models for fault diagnosis, evaluation and prediction based on support vector machine. Finally, realize the fault diagnosis system of wind generator gearbox based on support vector machine through dynamic link library throuth LabVIEW on sbRIO.