Research On The Condition Monitoring Of The Wind Turbine Based On The Cumulative Sum Charts

The operation and maintenance of wind farm is the key to further promotion and development of wind energy.In order to reduce the equipment damage,arrange the maintenance schedule reasonably and maintain the high efficiency of the wind farm,it is more and more important to study the condition inspection of the wind turbine.The research shows that the wind wheel is a key part of the unit,and the maintenance cost is high.Therefore,this paper monitors the operation status of the whole wind turbine and the wind wheel.In theory,because the Cumulative sum chart considers the data of the historical observation point,the detection effect of the small drift with trend is obviously better than the Shewhart control chart.Cumulative sum charts is the result of the further development of the Shewhart control chart.A state monitoring method for wind turbines based on Cumulative sum chart is proposed in this paper.For the research of the whole state monitoring of the wind turbine,the Bin method is used to get the power characteristics of the unit,and then the output power is monitored by the Cummulative sum chart,thus the state monitoring of the whole machine is realized.For the research of wind turbine operation,firstly,robust least squares regression is applied to establish the prediction model of hub speed under normal operation mode.Then,we use Lilliefors test to test the residuals in nonnal distribution,and further get the estimation of related parameters.At last,the normalized residual is used as the test statistic,and the Cumulative sum chart based on the residual is set up to monitor the running state of the wind wheel,thus the prediction of the fault is realized.Finally,this paper analyzes the actual operation monitoring data of a wind farm in Xinhua,Hunan,and compares it with the Shewhart control chart.The empirical analysis results show that the Cumulative sum chart based on SCADA data is an improvement of the Shewhart chart,which can realize the detection of the wind turbine operating status,discover the abnormality of the whole wind turbine and the wind wheel,and predict potential faults in advance.Through this research,we can speed up the process of maintenance,avoid the accident of equipment damage,and help the wind farm to run safely and efficiently.