基于改进GSP的电力通信告警关联挖掘研究

Research on Alarm Association Mining in Electric Power Communication Based on Improved GSP

随着电力通信网络规模不断扩大,网架结构日益复杂,其网络告警产生的原因与机理也多种多样。面对繁多复杂的通信告警,需要减少对人工专家知识的依赖,提高故障定位的准确性和效率。论文分析了主流通信告警关联分析方法存在的不足,针对电力通信网络设备网元数量多、告警数据数量大等特点,将序列模式挖掘与网络拓扑约束相结合,有效提高了电力通信告警关联分析的效率和精度。通过算例分析,该方法对电力通信告警关联分析具有较好的适用性,对提升通信运维管理水平具有一定的实际意义。

With the expanding scale of power communication network,the network structure is becoming more and more com⁃plex,and the causes and mechanisms of network alarm are also diverse.Faced with a variety of complex communication alarms,it is necessary to reduce the dependence on artificial expert knowledge and improve the accuracy and efficiency of fault location.This paper analyses the shortcomings of the main communication alarm correlation analysis methods.In view of the characteristics of large number of network elements and alarm data in power communication network equipment,the combination of sequential pattern mining and network topology constraints effectively improves the efficiency and accuracy of the power communication alarm correla⁃tion analysis.Through the example analysis,this method has good applicability to the correlation analysis of electric power communi⁃cation alarm,and has certain practical significance to improve the level of communication operation and maintenance management.