大数据环境下的输变电设备监控信息异常检测研究

Research on Abnormal Detection of Monitoring Information of Power Transmission andTransformation Equipment in the Big Data Environment

为了有效应对非线性关系给输变电设备监控信息异常检测带来的挑战,在大数据背景下,提出基于聚类分离信息特征的输变电设备监控信息异常检测方法。采用聚类分离信息特征的方法,聚类分离信息的时域特征、频域特征以及空间特征,并将输变电设备的监控数据进行特征化处理。建立一阶拟合模型拟合处理输变电设备的状态参数,关联聚类分离得到的信息特征与状态参数。根据特征和状态参数之间的权重值,实现输变电设备监控信息异常检测。实验结果表明,应用所提方法得到的异常点离群概率低于0.103,满足了输变电设备监控信息异常检测的需求。

In order to effectively address the challenges posed by non-linear relationships in detecting abnormal monitoring information of power transmission and transformation equipment,a method for detecting abnormal monitoring information of power transmission and transformation equipment based on clustering and separation of information features is proposed in the context of big data.Adopting the method of clustering to separate information features,clustering separates the time-domain,frequency-domain,and spatial features of information,and characterizes the monitoring data of power transmission and transformation equipment.Establish a first-order fitting model to fit and process the state parameters of power transmission and transformation equipment,and associate the information features and state parameters obtained through clustering and separation.Implement abnormal detection of monitoring information for power transmission and transformation equipment based on the weight values between features and state parameters.The experimental results show that the outlier probability obtained by applying the proposed method is less than 0.103,which meets the requirements of anomaly detection in monitoring information of power transmission and transformation equipment.