微型电脑应用2026，Vol.42Issue(1)：26-29,4.
离散化思维下异常用电行为数据检测方法设计
Design of Data Detection Method for Abnormal Electricity Consumption Behavior under Discrete Thinking
摘要
Abstract
Existing user behavior analysis methods mainly focus on integrating data,which have the shortcomings of strong sub-jectivity and physical models that are difficult to cope with the randomness and uncertainty of user behavior in complex power grid environments.Bad load data may appear,and the existing bad data may affect the accurate analysis of user behavior.Therefore,an automatic detection method for abnormal electricity consumption behavior data is proposed under discrete think-ing.By discretizing user electricity load data,continuous data can be transformed into discrete data,and the similarity between data can be compared to identify and handle issues such as outliers and missing values,thereby obtaining complete user electric-ity load data.Abnormal behavior features are mined from complete data to construct an automatic detection model.The pro-posed model identifies and detects potential abnormal behavior by training data on known normal and abnormal electricity con-sumption behavior.The experimental results show that using the proposed method for automatic detection of abnormal electric-ity consumption behavior has high detection accuracy and efficiency,and has broad application prospects in the detection of ab-normal electricity consumption behavior data in distribution networks.关键词
异常用电行为/特征挖掘/自动检测方法/数据清洗/检测模型Key words
abnormal electricity consumption behavior/feature mining/automatic detection method/data cleaning/detection model分类
信息技术与安全科学引用本文复制引用
冯维元,于雪辉,赖松乐,杨扬,熊会超..离散化思维下异常用电行为数据检测方法设计[J].微型电脑应用,2026,42(1):26-29,4.基金项目
河南省自然科学基金项目资助(192800310126) （192800310126）
