摘要
为提高风电机组部件故障预警的精度和速度,文章提出了一种基于SCADA数据的风电机组部件故障预警方法,用于解决现存的风电机组部件故障预警时间与故障预警精度的矛盾。首先SCADA历史数据通过数据预处理与BP神经网络建立机组部件正常状态模型,随后以该模型为基础结合基于马氏距离的数据统计方法形成故障预警判据,并将结果反馈给SCADA系统达到风电机组部件故障预警目的。仿真结果表明,使用该方法能够提前2个月识别主轴承故障信号,同时该方法能发现SCADA系统误报故障。
In order to improve the precision and speed of early warning of wind turbine components failure. In this paper, a fault warning method for the components of a wind turbine based on SCADA data is proposed. To solve the contradiction between the fault warning time of the existing wind turbine components and the accuracy of the fault warning, First SCADA historical data set up the normal state model of unit components through data preprocessing and BP neural network. Then the fault warning criterion is formed by using the model based data statistics method based on Mahalanobis distance. The results are fed back to the SCADA system to achieve the purpose of fault warning for the components of the wind turbine. The simulation results show that this method can identify the fault signal of the main bearing 2 months ahead of time, and the method can find the fault of SCADA system misinformation.
作者
吴亚联
梁坤鑫
苏永新
詹俊
Wu Yalian;Liang Kunxin;Su Yongxin;Zhan Jun(College of Information Engineering,Xiangtan University,Xiangtan 411105,China;Hunan Ulitech Automation System Co.,Ltd.,Changsha 410205,China)
出处
《无线互联科技》
2018年第13期122-127,共6页
Wireless Internet Technology
基金
湖南省自然科学基金资助项目
项目编号:2015JJ5027
国家自然科学基金资助项目
项目编号:61379063
