Study On Fault Diagnosis Of Gearbox Based On Hidden Markov Model

With the development of science and technology, the issue of dynamic process has attracted much attention in large machinery equipment such as wind turbines. The sudden failure of Machinery will increase its maintenance burden and production cost, and the reducing of production efficiency will directly impair the economic earnings of enterprises. As an industry with high degree of risk, to ensure safe operation of the wind turbine, it is necessary for wind power industry to take effective measures on their condition monitoring and maintenance of critical equipment. This paper mainly research on the means of wind turbine gearbox fault diagnosis, Due to the gearbox has the maximum impact on wind turbine gearbox. Based on the dynamic environment of wind turbines in operation and starting from the analysis of vibration signals, we apply the method of fault diagnosis to the gearbox state observation, modeling and evaluating. We will find the signs of the failure in the early period through online monitoring system of gearbox, so that to take early measures to maintain or replace the failure gearbox to avoid greater losses.Gearbox failures are important issues that affect the safe operation of the wind turbine, so it has a great significance to take methods to identify the operational status timely for the safe operation of gearbox. The traditional method of fault diagnosis has a lower accuracy and reliability, more and more intelligent diagnosis method was introduced to fault diagnosis. In this paper, Hidden Markov Model is applied to identify the failure modes of wind turbine gearbox, contain three parts:1) Study the methods of feature extraction from vibration signal, and optimized selection according to the sensitivity of the feature values; 2) Identifying different vibration signals of gearbox bearing with the application of HMM, then analyzing the result of diagnosis; 3) Make a preliminary study on the prediction method of remaining useful life on the basis of fault diagnosis. The results showed that:HMM can quickly and effectively identify the state of the gearbox such as wear, tooth breaking and some other failure modes, it has a good applicability and can be used in fault diagnosis in prediction of actual gearbox system.