Research On The Early Fault Diagnosis Method Of Wind Turbine Gearbox

Gearbox is one of the high failure rate components of the wind turbine. Because of its small installation location, maintenance difficulties and high maintenance costs, condition monitoring and fault diagnosis for wind turbine gearbox is more practical and economic. In this paper, wind turbine gearbox is the research object. Aim at the time-varying characteristic and non-stationary of its vibration signal, starting from the failure mechanism and fault characteristic, the early fault diagnosis method of wind turbine gearbox is in-depth researched.Based on the analysis of failure mechanism and vibration mechanism of wind turbine gearbox, the combined with failure modes of wind turbine gearbox components, the typical failure and vibration signal features of wind turbine gearbox are summarized by simulation.A new wind turbine gearbox fault diagnosis method combining with singular value decomposition noise reduction method and order analysis is proposed. The non-stationary signal processing method is studied, and three methods for instantaneous frequency estimation are introduced. The Hilbert-Huang Transform and central finite difference method are combined for the instantaneous frequency estimation. The basic principle of singular value decomposition noise reduction method is introduced. The selection principle of embedding dimension and delay time for phase space reconstruction are summarized. The principle of order analysis order analysis is discussed. The implementation steps of calculation order analysis method and the order analysis method based on instantaneous frequency estimation are summarized. The order analysis combining with singular value decomposition noise reduction method and instantaneous frequency estimation is proposed. The results are verified by the simulation signal.A wind turbine gearbox fault diagnosis experiment table is built. With the fault simulation experiment and measured signal, the early fault diagnosis method of wind turbine gearbox based on SVD Noise Reduction and order analysis is verified to be correct and effective.The subsystem of remote condition monitoring and fault diagnosis system for wind turbine gearbox named data analysis and diagnosis system is developed. The scheme of remote condition monitoring and fault diagnosis system for wind turbine gearbox is designed. With the use of modern signal processing technology, network programming, wireless sensor technology, virtual instrument technology, database technology, the subsystem is developed. Through the analysis and diagnosis of practical wind turbine gearbox vibration signal, the subsystem is verified to be reliable and effective.