基于K-SVD与FDTW的电力负荷数据修复研究  

Research on Power Load Data Restoration Based on K-SVD and FDTW

作　　者：姚晓明 YAO Xiaoming(State Grid Beijing Electric Power Company,Beijing 100031,China)

机构地区：[1]国网北京市电力公司,北京100031

出　　处：《自动化应用》2025年第4期92-94,共3页Automation Application

摘　　要：在电力负荷数据修复领域,现有方法在处理数据缺失和异常值时存在一定的局限性。基于K-SVD和FDTW两种算法,提出了一种新的电力负荷数据修复方法。旨在通过结合K-SVD算法的稀疏编码特性和FDTW算法的高效时间序列匹配能力,开发出一种能够提高数据修复精度和计算效率的混合算法,为电力系统的数据完整性和准确性提供可靠保障。In the field of power load data restoration,the existing methods have some limitations in dealing with missing data and outliers.This paper proposes a new approach to power load data restoration based on two algorithms,K-SVD and FDTW.It aims to develop a hybrid algorithm that can improve the accuracy and computational efficiency of data restoration by combining the sparse coding characteristics of the K-SVD algorithm and the efficient time-series matching capability of the FDTW algorithm,so as to provide a reliable guarantee for the data integrity and accuracy of the power system.

关 键 词：K-SVD FDTW 电力负荷 数据修复 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP391[自动化与计算机技术—控制科学与工程]