基于模糊规则的电力系统运行数据异常识别研究  

Fuzzy-rule-based Identification of Operating Data Anomalies of Power Systems

作　　者：石贤伦 刘凯 任凯 SHI Xianlun;LIU Kai;REN Kai(State Grid Ningbo Zhenhai District Power Supply Company,Ningbo 315000,China)

机构地区：[1]国网宁波市镇海区供电公司,浙江宁波315000

出　　处：《电工技术》2024年第1期60-62,共3页Electric Engineering

摘　　要：在识别电力系统运行数据异常的过程中,由于缺乏对数据之间关联关系的分析,识别结果的误差相对较大,因此提出一种基于模糊规则的电力系统运行数据异常识别方法。在构建电力系统运行数据模糊规则阶段,充分考虑了电力系统运行数据自身的属性特征,以合理粒度为原则,构建了具有多层结构的模糊规则,对数据的特征参量进行分割处理后,根据数据集中隐含的结构信息构建了电力系统运行数据之间的关联关系。在数据异常识别阶段,将模糊规则作为判断标准,计算电力系统运行数据的特异性程度,并通过约束模糊粒度保障识别结果的精度。测试结果证明,设计方法对不同程度数据异常的识别误差始终稳定在3.0%以内。Current studies on identifying operating data anomalies in power systems are generally inadequate in analyzing inter-data correlations,leading to relatively large deviation of identification results.In view of this the present work studied a fuzzy-rule-based method to identify operating data anomalies in power systems.A multi-layer-structured fuzzy rule was constructed by fully considering inherent attribute characteristics of power system operating data and following the justifiable granularity principle.After segmenting characteristic parameters of the data,the inter-data correlations were determined according to hidden structural information in the dataset.In the stage of data anomalies identification,taking fuzzy rules as judging criteria,the specificity of operating data was calculated and the accuracy of identification was ameliorated by restrained fuzzy granularity.The proposed identification method was proved by test results capable of achieving identification results whose deviation remained always within 3.0%with respect to operating data with different extents of anomalies.

关 键 词：模糊规则 电力系统 数据异常 合理粒度 特征参量 特异性 模糊粒度 

分 类 号：TM73[电气工程—电力系统及自动化]