欠定分离机制下基于特征滤波的居民负荷非侵入辨识算法

发布时间：2018-01-19 04:21

  本文关键词： 非侵入负荷监测 负荷分解 负荷辨识 欠定求解 特征滤波　出处：《电力系统自动化》2017年20期 　论文类型：期刊论文

【摘要】：通过非侵入采集模式下电流信号的欠定求解实现了负荷分解,获取了各独立负荷的完整电流,在负荷分解基础上实现了状态辨识。利用居民用户的负荷操作习惯将难以求解的欠定问题优化建模,转化为一维欠定问题,将求解模型建立为单位时间间隔仅从采集信号中分离两路信号。依据电流频域信号的稀疏性通过两步迭代收缩阈值算法得到最优解,使每个投入运行的负荷均可独立分解。通过先验方式获取用电网络各负荷的特征电流形成特征滤波器组,对分解电流进行频域滤波,通过对滤波后频率分量的量化判决实现负荷辨识。利用实际采集的用电数据验证了算法的有效性,能够有效实现负荷分解,并准确判断负荷状态。
[Abstract]:The load decomposition is realized by solving the underdetermined current signal in the non-invasive acquisition mode, and the complete current of each independent load is obtained. On the basis of load decomposition, the state identification is realized. By using the load operation habits of resident users, the difficult underdetermined problem is modeled and transformed into one-dimensional underdetermined problem. The solution model is established as a unit time interval to separate only two signals from the collected signal, and the optimal solution is obtained by two-step iterative shrinkage threshold algorithm according to the sparsity of the current frequency domain signal. Each load in operation can be decomposed independently. The characteristic current of each load of power network is obtained by a prior method to form a characteristic filter bank, and the decomposition current is filtered in frequency domain. The load identification is realized by quantifying the frequency component after filtering, and the validity of the algorithm is verified by the actual electric data collected. The load decomposition can be realized effectively and the load state can be judged accurately.
【作者单位】： 华北电力大学电气与电子工程学院;国网江苏省电力公司电力科学研究院;
【基金】：中央高校基本科研业务费专项资金资助项目(2016MS13) 国家重点研发计划资助项目(2016YFB0901104)~~
【分类号】：TM714
【正文快照】： 上网日期:2017-08-08。0引言智能电网的发展使传统电力工业向高度集约化、知识化、技术化方向转变,需求侧管理的重要性日益凸显[1]。通过需求侧管理优化低压用户端,实现有效的负荷管理,能够在合理有效用电的基础上减少电量消耗和电力需求,提高电能利用效率[2]。实现用电的全面，

本文编号：1442609