| 引用本文: | 黄剑湘,林 铮,刘可真,等.考虑换流站海量事件的关联规则挖掘分析方法[J].电力系统保护与控制,2022,50(12):117-126. |
| HUANG Jianxiang,LIN Zheng,LIU Kezhen,et al.Association rule mining analysis method considering massive events in a converter station[J].Power System Protection and Control,2022,50(12):117-126 |
|
| 摘要: |
| 为提高换流站运维人员面对海量生成事件的分析能力，提出一种考虑换流站海量事件的关联规则挖掘分析方法。首先，利用原始事件元组特性进行记录事件与响应日志的实体特征筛选，并进行换流站实体特征的布尔映射与关联挖掘建模。然后，利用互信息(MI)原理与对称不确定性(SU)理论改进FP-Growth算法。最后，基于改进算法进行换流站事件关联分析，进而基于关联规则结果进行换流站异常反馈。通过挖掘昆柳龙直流换流站调试期间海量生成事件，表明所提出的方法可以有效地从海量事件中提取判断特征与结果特征的强关联规则，及时发现换流站的设备异常动作，并为运维分析提供决策支撑。 |
| 关键词: 事件元组特性 布尔映射 改进FP-Growth算法 异常反馈 昆柳龙直流换流站 |
| DOI：DOI: 10.19783/j.cnki.pspc.211148 |
| 分类号: |
| 基金项目:国家自然科学基金项目资助(51907084)；中国南方电网有限责任公司超高压输电公司核心攻关科技项目资助(CGYKJXM20180212)；云南省应用基础研究计划项目资助(202101AT070080) |
|
| Association rule mining analysis method considering massive events in a converter station |
|
HUANG Jianxiang1, LIN Zheng2, LIU Kezhen2, LUO Zhao2, YU Jinyun1, XU Feng1
|
|
(1. Kunming Bureau of CSG EHV Transmission Company, Kunming 650217, China;
2. Faculty of Electric Power Engineering, Kunming University of Science and Technology, Kunming 650500, China)
|
| Abstract: |
| To improve the ability of converter station operational and maintenance personnel to analyze massive generated events, this paper proposes an association rule mining analysis method considering a large number of events in converter stations. First, the entity features of recorded event and response logs are filtered using the original event tuple features. Then Boolean mapping and association mining modeling of the entity features of the converter station are performed. Then, an FP-Growth algorithm is improved using the mutual information (MI) principle and symmetric uncertainty (SU) theory. Finally, based on the improved algorithm, event correlation analysis of the converter station is carried out, and then the feedback of the converter station anomaly is carried out based on the results of the correlation rules. By mining the massive generated events during the commissioning of the Kun-Liu-Long DC converter station, it is shown that the proposed method can effectively extract strong correlation rules of judgment and result features from the massive events, discover the abnormal equipment actions of the converter station in time, and provide the decision support for operation and maintenance analysis.
This work is supported by the National Natural Science Foundation of China (No. 51907084). |
| Key words: event tuple characteristics Boolean mapping improved FP-Growth algorithm abnormality feedback Kun-Liu-Long DC converter station |
