Study On Vibration Fault Diagnosis Of Wind Turbine Gear Box Based On Multidimensional Permutation Entropy And Support Vector Machine

With the vigorous development of new energy resources cause in our country,the proportion of wind power are increasing. After years of development, however,along with the operation of the accumulation of wind turbines use fixed number of year, there were many wind farms have appeared unit key parts of the fault. A lot of wind turbines, life do not accord with the design life. There were also many problems with high failure rate, maintenance cycle is long, or trouble-free operation time is short. Especially in the early days of commissioning and still occupy a larger proportion in wind power of doubly-fed wind power units in the above picture is more obvious, and the core components such fan gear box failure rate as high as 40% - 50%.Therefore, it is more important to carry out the research on the fault diagnosis of wind turbine gearboxIn view of this, This paper studies the gearbox of wind turbine, and study the vibration fault diagnosis.The main content is as follows:I. Permutation entropy algorithm is introduced to early failure analysis of gear box vibration signal, and then introduce multi-scale permutation entropy fan gearbox vibration signal feature extraction algorithm, for the wind turbine vibration fault diagnosis of gear box opened up a new research ideas and solutions.2. A fault diagnosis model based on genetic algorithm optimization support vector machine was constructed. In this paper, on the basis of support vector machine(SVM) is used for fault diagnosis, the width of the radial basis kernel f'unction parameter and error penalty parameter discussed the selection of Ithe two important parameters, considering the demand of the other samples exist large, limitations of strong and weak generalization ability and premature problem, eventually determine the parameters of support vector machine was optimized by genetic algorithm, and then constructs a model of fault diagnosis classification effect is ideal.3. Paper is proposed based on multiscale permutation entropy and support vector machine (SVM) of the wind turbine gearbox fault diagnosis methods, validated by the instance data and analysis, the results show that the method has high diagnostic accuracy and diagnostic accuracy of 93.333% on average.The diagnosis of the above research results show that the proposed method can more accurately the types of vibration fault diagnosis of wind turbines of wind power gear box fault diagnosis research provides a new train of thought...