基于决策树的电力通信网全链路数据异常检测方法

Decision tree-based full-link data anomaly detection method for power communication network

数据异常会造成电力通信网全链路故障,为建立更加稳定的电网通信环境,提出基于决策树的电力通信网全链路数据异常检测方法。建立决策树组织,通过剪枝处理实现对电力通信网异常数据的挖掘。根据数据标准化原则,定义链路标签,联合相关电力通信数据,完成电力通信网全链路数据异常检测。实验结果表明,所提方法的异常传输行为的数据总量,在电力通信数据样本中所占比例低于13.9%。表明所提方法能够维护电网通信稳定性,解决了电力通信网全链路故障问题。

Data anomaly will cause full-link failure of power communication network.In order to establish a more stable power network communication environment,a decision tree-based full-link data anomaly detection method for power communication network is proposed.Establish the decision tree organization,and realize the mining of abnormal data of electric power communication network through pruning.According to the principle of data standardization,the link label is defined,and the anomaly detection of the whole link data of the power communication network is completed by combining the relevant power communication data.The experimental results show that the total amount of abnormal transmission behavior data of the proposed method accounts for less than 13.9%of the power communication data samples.The results show that the proposed method can maintain the stability of power network communication and solve the problem of full-link fault of power communication network.