| 基于时间概率分布与电器组合超状态的非侵入式负荷分解 |
| |
| 引用本文： | 燕续峰,翟少鹏,何光宇.基于时间概率分布与电器组合超状态的非侵入式负荷分解[J].电力系统自动化,2018,42(24):53-58. |
| |
| 作者姓名： | 燕续峰 翟少鹏 何光宇 |
| |
| 作者单位： | 上海交通大学电子信息与电气工程学院, 上海市 200240,上海交通大学电子信息与电气工程学院, 上海市 200240,上海交通大学电子信息与电气工程学院, 上海市 200240 |
| |
| 基金项目： | 国家自然科学基金资助项目(51877134) |
| |
| 摘 要： | 非侵入式负荷监测是智能用电网络中的重要技术，为研究非侵入监测下电器的用电情况，提出了一种基于电器时间概率分布和电器组合超状态匹配的负荷分解方法。首先对电器的功率数据进行状态提取，并利用电器运行的时间信息提取每个状态的时间概率分布。根据电器的工作状态组合，构建超状态，利用家庭历史用电数据缩减超状态空间，并针对超状态功率重叠问题对其进行聚类得到S超状态。在负荷分解阶段利用S超状态匹配方法对非侵入式数据分段，并利用时间概率最大似然估计分解结果。最后，通过数据集分解结果验证了该算法的有效性。
|
| 关 键 词： | 非侵入式负荷分解 时间概率分布 超状态 S超状态匹配 |
| 收稿时间： | 2018/3/22 0:00:00 |
| 修稿时间： | 2018/9/25 0:00:00 |
Non-intrusive Load Disaggregation Based on Time Probability Distribution and Super State of Appliances |
| |
| YAN Xufeng,ZHAI Shaopeng and HE Guangyu.Non-intrusive Load Disaggregation Based on Time Probability Distribution and Super State of Appliances[J].Automation of Electric Power Systems,2018,42(24):53-58. |
| |
| Authors: | YAN Xufeng ZHAI Shaopeng and HE Guangyu |
| |
| Affiliation: | School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China,School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China and School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China |
| |
| Abstract: | Non-intrusive load monitoring technique plays an important role in the intelligent electric appliance network. In order to analyze the power consumption of household appliances under non-intrusive load monitoring, a load disaggregation algorithm based on time probability distribution of appliances and super states matching is proposed. Firstly, the state of each appliance is extracted based on its power data and the time probability distribution of each state is extracted using the time information. The super states are constructed based on the state combinations of electrical appliances. The space of super states is reduced by considering households historical data. The S-super states are obtained by clustering super states to reduce the impact of power overlap. At the load disaggregation stage, non-intrusive power data is segmented by using the S-state matching method. The optimal solution is found by using the maximum likelihood of time probability. Finally, the effectiveness of the algorithm is verified by dataset test results. |
| |
| Keywords: | non-intrusive load disaggregation time probability distribution super state S-super state matching |
|
| 点击此处可从《电力系统自动化》浏览原始摘要信息 |
|
点击此处可从《电力系统自动化》下载全文 |
