Research On Fault Diagnosis Of Wind Turbine Gearbox Based On Variational Mode Decomposition

Since the end of the 20^th century,the wind power industry has developed rapidly throughout the world,and the installed capacity of wind turbines has a sustained increase.The stability and safety of wind turbine systems have ushered in new challenges.Under the circumstances,condition monitoring and fault diagnosis technology of wind turbines have become a pivotal issue for ensuring a brighter development of the wind power industry.Gearbox is a central component of the transmission system of wind turbine,monitoring and analyzing of the running state are of great significance to the operation and maintenance of the whole wind turbine.In this paper,wind turbine gearbox is taken as the research object,aiming at the problem of vibration signal feature extraction in gearbox fault diagnosis.Based on Variational Mode Decomposition?VMD?,in terms of weak feature extraction and nonlinear feature quantitative analysis,the theoretical research and experimental verification of fault diagnosis method are implemented,which propose valid methods for wind turbine gearbox condition monitoring and fault diagnosis.The mian research works are as follows:Firstly,the wind turbine gearbox structure characteristics,failure mechanism and vibration fault characteristics are analyzed.Moreover,aiming at the issue that the gearbox fault vibration signal is weak,non-linear and difficult to extract and quantify,the feature extraction and quantitative analysis methods of vibration signal feature based on VMD are studied to realize the valid extraction and exact description of fault feature of wind turbine gearbox.Secondly,the time-frequency analysis method of vibration signal based on VMD and Resonance-Based Spare Signal Decomposition?RB-SSD?is used to enhance and extract the weak feature of the vibration signal interference with backgroud noise of gearbox bearing under different fault conditions.Simulation analysis and bearing diagnosis experiment verify that the effectiveness of proposed method.Finally,in order to acquire the faint change in complicated fault signals of wind turbine gearbox and announce the characteristics of the change from the nonlinear dynamics,a new indicator named Variational Mode Multiscale Weighted-Permutation Entropy?VMMWPE?is proposed,which is based on VMD and Weighted-Permutation Entropy?WPE?.The VMMWPE can quantitatively analyze the different distribution of ordinal pattern corresponding to the nonlinear characters in high-dimensional phase space among several fault types.It's applied in the composite fault diagnosis of gearbox,and effectively represent nonlinear characteristics of gearbox fault signals.Particle Swarm Optimization Support Vector Machine?PSO-SVM?is used to further verify the effectiveness of the proposed indicator,and realize the fault pattern recognition.