摘要
采用温度趋势分析的方法监测风电机组发电机的运行状态;使用BP神经网络,建立发电机正常工作状态下的温度模型并用其进行温度预测;合理选择训练样本,使BP模型覆盖发电机的正常工作空间;当发电机工作异常时,其动态特性偏离正常工作空间,导致BP网络温度模型预测残差的分布特性发生变化;当残差超过预先设定的阈值时,发出报警信息,提示运行人员检查设备状态。
Temperature trend analysis is an effective method to monitor the wind turbine generator condition.BP neural network has been used to construct temperature model of generator under normal conditions and applied for temperature forecasting.BP model could cover the normal work space of generator based on reasonable selecting training samples.When the generator works abnormally,its dynamic characteristic deviates from the normal working space,which may cause the change of distribution characteristic of the forecasted residual by BP network model.When the residual exceeds a predefined threshold,alarm will be triggered to remind the operator to check the generator condition.
出处
《吉林电力》
2012年第5期29-32,共4页
Jilin Electric Power
关键词
状态监测
发电机
BP神经网络
残差
condition monitoring
generator
back propagation(BP) neural network
residuals