摘要
双碳目标的提出,促进了电网朝着清洁低碳和安全可控等方向发展,线损是节能减排的直接体现和重要技术指标,针对现有配电网线损异常检测方法存在的检测精度差和效率低等问题,在线损异常检测系统的基础上,提出了一种结合灰色关联分析、改进K-means算法和孤立森林算法的配电网线损异常检测方法。通过灰色关联分析和改进K-means算法优化孤立森林算法,灰色关联分析完成特征属性的筛选,改进K-means算法完成数据聚类处理,提高了孤立森林算法异常检测的准确率和效率。结果表明,所提方法与常规方法相比,在多项指标上具有最优的检验效果,检验精度达到100%,平均检测时间为0.0402 s,为双碳目标实现提供了一定的支持。
The proposal of the dual-carbon target promotes the development of the power grid in the direction of clean,low-carbon,safe and controllable,etc.Line loss is a direct reflection of energy saving and emission reduction and an important technical indicator.In view of the problems such as poor detection accuracy and low efficiency in the existing line loss anomaly detection methods,a distribution network line loss anomaly detection method combining grey correlation analysis,improved K-means algorithm,and isolated forest algorithm is proposed based on line loss anomaly detection system.The isolated forest algorithm is optimized through grey correlation analysis and improved K-means algorithm,the feature attribute screening is completed through grey correlation analysis,and data clustering processing is completed through improved K-means algorithm,and the accuracy and efficiency of anomaly detection in the isolated forest algorithm are improved.The results show that the proposed method has the best testing effect on multiple indicators compared to conventional methods,with a testing accuracy of 100%and an average detection time is 0.0402 seconds,providing certain support for achieving the carbon peaking and carbon neutrality goals.
作者
吴若冰
张振超
WU Ruobing;ZHANG Zhenchao(Information Center,Yunnan Power Grid Co.,Ltd.,Kunming 650000,China)
出处
《电测与仪表》
北大核心
2024年第9期145-150,共6页
Electrical Measurement & Instrumentation
基金
中国南方电网有限责任公司科技项目(YDKJ 23030064)。
关键词
配电网线损
异常检测
灰色关联分析
K-MEANS算法
孤立森林算法
双碳目标
distribution network line loss
abnormal detection
grey relation analysis
K-means algorithm
isolation forest algorithm
carbon peaking and carbon neutrality goal
