基于信号稀疏化欠定求解的居民用户非侵入式负荷分解算法

发布时间：2018-02-25 04:14

  本文关键词： 非侵入负荷分解 稀疏性 欠定求解 相似系数　出处：《电网技术》2017年09期 　论文类型：期刊论文

【摘要】：非侵入式负荷监测是实现智能用电的关键技术,负荷分解是实现该技术的重要环节。立足于非侵入采集模式下电流信号的欠定求解,研究了一种负荷分解算法。利用居民用户电器启动时间存在时间差的特点,将求解模型建立为单位时间间隔内仅从采集信号中分离两路信号,一路为新投入运行负荷的独立电流,一路为上一时刻的混合电流。该模型使每个投入运行的负荷均可独立分解,同时保证了欠定维度小,从而可有效求解。求解过程结合负荷电流信号的频域稀疏性,将欠定方程转化为最优化约束问题,通过两步迭代收缩阈值算法从混合信号中恢复出两路电流信号,并通过相似系数判断分离是否有效。利用实测用电数据验证了算法的有效性,通过正迭代方法得到当前参与运行的用电负荷,并根据相似系数确定负荷类型,实现了负荷辨识。
[Abstract]:Non-intrusive load monitoring is the key technology to realize intelligent power consumption, and load decomposition is an important link to realize this technology. In this paper, a load decomposition algorithm is studied. Based on the characteristic of time difference in the start-up time of household electrical appliances, the solution model is established to separate only two signals from the collected signals in unit time interval. An independent current for a new operational load and a mixed current for the last time. The model allows each load to be decomposed independently, while ensuring a small underdetermined dimension. The solution process combines the frequency domain sparsity of the load current signal, transforms the underdetermined equation into an optimization constraint problem, and recovers two current signals from the mixed signal by the two-step iterative shrinkage threshold algorithm. The validity of the algorithm is verified by the measured power consumption data, and the current load is obtained by the positive iteration method, and the load type is determined according to the similarity coefficient, and the load identification is realized.
【作者单位】： 华北电力大学电气与电子工程学院;
【基金】：国家重点研究发展计划项目(2016YFB0901104) 中央高校基本科研业务费专项资金资助项目(2016MS13)~~
【分类号】：TM714;TM73
，

本文编号：1532963