计算机与数字工程2026，Vol.54Issue(2)：444-449,6.DOI:10.3969/j.issn.1672-9722.2026.02.025
基于Spark和K-means++的电力数据异常检测
Power Data Anomaly Detection Based on Spark and K-means++
张磊 1余粟2
作者信息
- 1. 上海工程技术大学机械与汽车工程学院 上海 201620
- 2. 上海工程技术大学工程实训中心 上海 201620
- 折叠
摘要
Abstract
With the gradual improvement of power grid information level,power data shows a trend of quantification and com-plexity.Aiming at the problems of low accuracy and poor stability of traditional power big data anomaly detection methods,insuffi-cient computing resources in the context of massive data,and low efficiency of MapReduce engine in Hadoop architecture in the con-text of repeated iterations such as machine learning,a parallel K-means++algorithm based on Spark is proposed.The simulation re-sults show that the algorithm has a good effect on the detection of abnormal data in the power data scenario,and can effectively im-prove the accuracy and convergence of clustering results.关键词
电力负荷/Spark/K-means++/异常检测Key words
power load/Spark/K-means++/anomaly detection分类
信息技术与安全科学引用本文复制引用
张磊,余粟..基于Spark和K-means++的电力数据异常检测[J].计算机与数字工程,2026,54(2):444-449,6.
