Research On Early Fault Warning And Diagnosis Of Wind Turbine Gear

Wind power as a more mature renewable energy has been widely used, but as a result of the harsh environment, wind turbine has higher failure rate. Gearbox as the most important part of the wind turbine, plays a power transmission role, the health of gearbox is directly related to the wind turbine operation condition. Strengthening the gearbox condition monitoring and achieving early fault warning and diagnosis for a wind turbine are of great significance to reduce operating costs and maintenance costs. This paper does research and analysis from the two aspects of SCADA parameters and vibration parameters of wind turbine gearbox, to finish the early fault warning and diagnosis of wear and modulation, mainly includes the following aspects.Firstly, the paper analyzes the main structure and vibration mechanism of the wind turbine gearbox, for typical fault mode of gearbox, analyzes the symptoms, reasons and impacts of faults. Through FTA analysis and of the gearbox, establishing the complete fault knowledge base and laying the theoretical foundation for the next fault warning and diagnostic work.Secondly, in view of SCADA parameters, take the selected parameters and calculate the history coefficient matrix and real time coefficient matrix and their deviation. Through the deviation realizing the state warning of the gearbox within a period time. In view of the vibration signal, this paper adopts a quick and simple interpolation method, transforming the time domain signal into angle domain signal and realizing the resample of complete cycle. Extract the dimensionless characteristics from angle-domain signals, map the relevant dimensionless characteristic values to a three dimensional space by the means of feature fusion, calculate the distance of the characteristic values and realize the early fault warning through the abnormal distance. the HHT method is introduced into the early gearbox fault diagnosis and the angle domain signal getting above is handled in the EMD decomposition. For each IMF component extract corresponding order ratio spectrum and energy spectrum, combined with the feature of fault, analyze the order ratio spectrum by energy changes. Through the energy changes to realize accurately locating the fault position and warning the fault degree.At last, the method of early fault warning and diagnosis is applied to engineering. The placement of sensors is determined and the kinds of sensors are also determined to the drive system of wind turbine. A simulation based on the actual parameters of gearbox is realized, the failure mode is identified through the method in the preview chapter. At last a diagnosis report is given, which is also used to lead the maintenance of the gearbox.