Research On The Wind Turbine Drive-train System State Evaluation Based On Operational Condition

With the growth of the installed capacity of wind turbines and the running time increased, the complex and large-scale wind turbine unit has become the trend of development. In order to accurately grasp the operational status of wind turbines,improving power generation performance and operational reliability of the unit, lower unit operation and maintenance costs, a method of fault warning of the drive-train system of wind turbines based on data mining is proposed in this paper, using the massive historical operating data, which is accumulated over the years of wind farm.Through the fault prognosis can reduce the number of equipment failures effectively,avoid major faults, while pre-arranged reasonable maintenance plan, will help improve the reliability of the equipment, but also to lay the foundation for the realization of state-based maintenance. Wind Turbine Transmission System is a key wind energy conversion system, wherein the gearbox failure is one of the largest unit that caused downtime. Wind turbine gearbox particularity of its operating environment, located in altitude, it is not conducive to testing and maintenance, which requires wind turbine gearbox for online condition monitoring, in order to be able to keep abreast the operational status and potential malfunction of the gearbox. In this paper, the gearbox of wind turbine drive-train system is used as the research object.The main contents are as follows:Firstly, this paper introduces the current domestic and foreign status of the operation and maintenance of the wind turbine, and analysis the opportunities and challenges facing wind turbine in operation and maintenance. Then introduce the research progress and future trends for improving wind turbine reliability and lower operation and maintenance costs.Secondly, gives an overview of the wind turbine works, including the basic structure of the wind turbine, transmission common faults, the basic structure and the failure mechanism of the gearbox and the like. Then introduce the basic structure of the wind farm SCADA(supervisory control and data acquisition system) and data acquisition parameters. After that make a simple summary and comparison of fault warning methods, and proposed a method based on SCADA system data mining of wind turbine fault warning based on existing conditions of wind farmsIn this paper, use a method based on SCADA system data mining of wind turbine fault warning. Preprocess historical data of SCADA system, select the features associated with the gearbox operating parameters, using the method of cluster analysis to classify the normal operation of the data to construct the gearbox healthy model library. And then through collecting the real-time data in the field, using NEST(nonlinear state estimate technology) theory to predict the real-time data and obtainthe predicted value, then computing the residuals of predicted value and actual value.Statistics residuals by sliding window, set the fault alert starting rules, when the residual value exceeds the set threshold, the system gives the fault warnings, and can query the fault warning reasons, to provide references for fault eliminating defects.Finally by way of example of the Health Evaluation Based on the data mining model is verified, the results show that the method can detect potential faults of wind turbine gearbox promptly and effectively, also gives fault warning in advance.In this paper, the research results can be integrated in the original wind farm SCADA system, lay the foundation for wind turbine real-time online status evaluation and fault warning, and provide a reference for wind turbine condition based maintenance.