Research On Wind Turbine Gearbox Condition Monitoring And Fault Diagnosis System

The oil crisis of last century cause many people started looking for the clean and renewable resources which could instead of the fossil energy. In the21st century, renewable energy has achieved great development, especially the wind energy. The wind power always maintains the fastest growth for its own advantages, and it has become the core energy except petroleum and chemical energy. By the end of2011, the global wind power total installed capacity is238.4GW and it still in rapid development. Along with the rapid development of the wind power, how to ensure the stable operation of the wind generator and how to resolve the failure of the wind turbine are problems that we must resolve, and this is also why we choose this research topic.In this paper, we choose the gearbox as the research object, and we will propose a new method to build the condition monitoring and fault diagnosis system for the gearbox. Gearbox is the main transmission device of the wind turbine, and it is also the component that most prone to failure. According to former data statistics, the longest parking time is caused by the failure of gearbox. The gearbox fault also caused the greatest loss of generating capacity, so we choose the gearbox as the research object. The following is the main work of this paper:In order to better resolve the gearbox fault diagnosis problem, we proposed an improved intuitionistic fuzzy support vector machine algorithm to solve this problem. The improved algorithm can improve the efficiency of the algorithm though the pre-extraction support vector method, and improve the accuracy of the algorithm though redefine the fuzzy membership and intuition index. The test results though artificial data sample and UCI data sample show that the improved algorithm is more efficiency and accuracy.In this paper, we build a fault diagnosis model to resolve the gearbox fault diagnosis problem, and the improved algorithm is used in the fault diagnosis model. The test result that our method is correct and this method can properly resolve the gearbox diagnosis problem.We proposed a new system structure to build the gearbox condition monitoring and fault diagnosis system. This system structure can make the condition monitoring system and fault diagnosis system work together. It is very useful and practical.