基于Apriori算法的气象业务监控系统告警关联分析

Alarm Correlation Analysis of Meteorological Business Monitoring System Based on Apriori Algorithm

基于气象信息中心传输业务、硬件设备监控系统中产生的告警信息未能准确定位故障根源及处理效率低的问题,通过数据接口调用和爬虫方法对监控系统中产生的历史告警信息进行收集,利用数据清洗方法做预处理、数据转换操作,引入Apriori关联分析算法对告警进行关联分析,利用提升度、置信度加权平均、置信度不平衡比等验证方法证实关联性结果的可信度,通过业务角度分析关联性分析的结果确定告警处理的优先级,依据告警优先级提高监控系统的故障处理效率。

Based on the problems of inaccurate identification of fault sources and low processing efficiency of alarm information generated by the transmission business of the Meteorological Information Center and the hardware equipment monitoring system,the historical alarm information generated by the monitoring system is collected through data interface calling and crawler methods.Data cleaning methods are used for preprocessing and data conversion operations.The Apriori association analysis algorithm is introduced to perform association analysis on alarms.The credibility of the association results is verified by using methods such as lift,weighted average of confidence,and confidence imbalance ratio.The priority of alarm processing is determined through business-oriented analysis of the association analysis results,and the efficiency of fault processing of the monitoring system is improved through alarm priority.