住宅用电非侵入式电力负荷监测与识别  

Non-intrusive power load monitoring and identification for residential power consumption

作　　者：陈飞 杨超[1] CHEN Fei;YANG Chao(School of Electrical Engineering,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学电气工程学院,贵阳550025

出　　处：《智能计算机与应用》2021年第9期75-84,共10页Intelligent Computer and Applications

基　　金：贵州省科学技术基金(黔科合基础[2019]1100)。

摘　　要：在住宅用电的非侵入式负荷监测与识别过程中,对负荷的投切时间检测与负荷类型的准确识别是重要的研究部分,本文提出了一种住宅用电非侵入式电力负荷监测与识别算法。首先,提出了一种基于短期环比变点检测算法,能够准确的检测出负荷的投切时间,进而对负荷的稳态特征进行高精度的提取;其次,利用融合莱维飞行与黄金正弦的蚁狮优化(LGSALO)支持向量机算法,实现家用电力负荷识别。实验结果表明,本文所提出的变点检测算法和负荷识别方法在实际应用过程中具备较高的准确率。In the process of non-invasive load monitoring and identification of residential power consumption,the detection of load switching time and accurate identification of load type is an important part of research. This paper proposes a non-invasive power load monitoring and identification algorithm for residential power consumption. Firstly,a change point detection algorithm based on short-term link comparison is proposed,which can accurately detect the load switching time. Then,the steady-state characteristics of the load is extracted with high precision,and the ant lion optimization(LGSALO)support vector machine algorithm is used to realize the household power load identification. The experimental results show that the proposed change point detection algorithm and load identification method have high accuracy in the practical application process.

关 键 词：非侵入式负荷监测 变点检测 负荷识别 短期环比 LGSALO 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]