基于FP-Growth算法的计量主站告警分析研究

Analysis and Research on Alarm of Master MeteringStation Based on FP-Growth Algorithm

随着电力企业信息技术的广泛应用,计量主站系统产生数量庞大和种类繁多的告警信息,缺乏多个告警事件之间的关联分析方法,因而造成误报、漏报等问题。基于计量系统中的历史告警数据,通过引入库尔钦斯基度量以及不平衡比,改进FP-Growth算法,排除关联规则挖掘过程中无意义关联规则的影响,确定计量主站告警的识别方法。结果表明:FP-Growth算法挖掘计量主站告警的关联规则准确率达到了91.2%,算法准确度较高;构建的FP-Growth算法挖掘计量主站告警关联规则具有较高的实际应用价值,剔除外部干扰因素的影响,协助电网公司更加精准地发现计量主站告警源,高效排查并解决计量主站告警,保障电网稳定运行。

With the extensive application of information technology in electric power enterprises,the master metering station system produces a large number of alarm information of various types,and lacks the correlation analysis method among multiple alarm events,thus causing problems such as false alarm and missing alarm.Based on the historical alarm data in the measurement system,the FP-Growth algorithm was improved through Kulczynski measurement and imbalance ratio,and the influence of meaningless associated rules in the discovering process of associated rules was eliminated to determine the identification method of master metering station alarming.The experimental results show that the accuracy of FP-Growth algorithm for discovering the associated alarm rules of the master metering station reaches 91.2%,i.e.,this algorithm accuracy is high;the constructed FP-Growth algorithm is of high practical application value on searching the associated rules of master metering station alarming,eliminating the influence of external interference factors,assisting power grid companies to more accurately discover the alarm sources of the master metering station,and efficiently troubleshooting and resolving the alarms of the master metering station,so as to ensure the stable operation of the power grid.