基于多特征融合的变电站运行异常预警研究

Research on Abnormal Warning of Substation Operation Based on Multi Feature Fusion

常规的变电站的运行异常预警研究中,没有对于运行状态趋势的数据源进行分析,导致预警中的样本数据不足,不能够在算法中减少时间消耗,达到快速预警。因此,提出基于多特征融合的变电站运行异常预警研究。通过离散时间序列对变电站运行状态趋势预警数据源进行分析,在变电站运行异常数据特征提取的基础上完成基于多特征融合的变电站异常状况模型建立,最后通过训练样本实现变电站运行异常预警研究。实验中实验组在220kV高压室中出现紧急程度的异常状况时,需要花费9s进行异常数据的侦查及信息传递,而对照组在同等情况下需要花费12s的时间,实验组比对照组对于异常状况的信息传递时间快了3s,结果可以表明基于多特征融合的异常预警研究消耗时间按更短,更有利于预警信息传递。

In the conventional research on the early warning of abnormal operation of substations,the data source of operation state trend is not analyzed,which leads to insufficient sample data in the early warning,which can not reduce the time consumption in the algorithm and achieve rapid early warning.Therefore,the research on early warning of abnormal operation of substation based on multi-feature fusion is proposed.The data source of early warning of substation operation trend is analyzed by discrete time series,and the abnormal situation model of substation based on multi-feature and multi-feature fusion is established on the basis of feature extraction of abnormal data of substation operation.Finally,the early warning research of abnormal operation of substation is realized through training samples.In the experiment,when there is an emergency abnormal situation in the 220kV high-voltage room,the experimental group needs to spend 9 seconds to detect abnormal data and transmit information,while the control group needs to spend 12 seconds in the same situation,and the information transmission time of the experimental group is 3 seconds faster than that of the control group.The results show that the abnormal early warning research based on multi-feature fusion takes less time and is more conducive to the transmission of early warning information.