摘要
由于风电机组的运行条件恶劣,在运行过程中经常会出现许多不确定的外界因素,这些因素使得风电机组各部件的故障率较高。采用小波BP神经网络的时间序列方法对风电机组的齿轮箱温度进行预测,并利用滑动窗口技术对其预测残差进行统计分析,然后通过分析齿轮箱温度的残差均值和标准差来预测齿轮箱温度是否存在异常情况或是故障隐患,从而达到预警目的。
The wind turbines are operated in harsh environment and always appear many uncertain outside factors,therefore all parts of the wind turbines have higher failure rate.In this paper,the time series method of wavelet BP neural network is used to building the gearbox temperature modeling,and then the sliding window technology is used to statistical analysis of the prediction residual.Finally,through the analysis of the gearbox temperature residual mean and standard deviation the prewarning information is gotten.The operator not only can see the real-time operating state,also can discover the abnormal situation and fault hidden trouble in time through this method,so they can prepare maintenance in advance and avoid the equipment damage.
出处
《电子测量与仪器学报》
CSCD
2012年第3期197-201,共5页
Journal of Electronic Measurement and Instrumentation