基于人工神经网络的水轮发电机组振动预测研究

Research on vibration prediction of hydroelectric generating units based on artificial neural network

为了解决目前水轮发电机组状态监测过程中振动报警数值设置单一,不能满足实际运行的问题,根据三峡电站升水位试验数据,提出了基于BP算法的水轮发电机组监测部件振动人工神经网络预测模型,并给出了具体算法。研究表明,由于建立在翔实的数据基础之上,该模型能够有效预测水轮发电机组监测部件的振动,通过设置合理的报警阈值空间,能够有效地减少机组误报警,提高报警的有效性。

In order to solve the problem that a single vibration alarm value can not satisfy the actual operation in the monitoring process of current hydroelectric generating units.Based on the experimental data of the rising-water test of TGP,we propose the prediction model of artificial neural network for vibration of hydroelectric generating units monitoring parts based on BP algorithm,and give the specific algorithm.The results show that the model can effectively predict the vibration of hydroelectric generating units on the basis of detailed data,and false alarms of the generating units can be reduced availably to improve the effectiveness of alarm by setting the reasonable alarm threshold space.