[1]张顺淼、陈铭龙.基于注意力机制与ConvBiLSTM的非侵入式负荷分解[J].福建工程学院学报,2020,18(04):336-342.[doi:10.3969/j.issn.1672-4348.2020.04.006]
　ZHANG Shunmiao,CHEN Minglong.Non-intrusive load decomposition based on attention mechanism and ConvBiLSTM[J].Journal of FuJian University of Technology,2020,18(04):336-342.[doi:10.3969/j.issn.1672-4348.2020.04.006]
点击复制

基于注意力机制与ConvBiLSTM的非侵入式负荷分解()
分享到：

《福建工程学院学报》[ISSN:2097-3853/CN:35-1351/Z]

卷:
第18卷
期数:
2020年04期
页码:
336-342
栏目:
出版日期:
2020-08-25

文章信息/Info

Title:
Non-intrusive load decomposition based on attention mechanism and ConvBiLSTM
作者:
张顺淼、陈铭龙
福建工程学院信息科学与工程学院
Author(s):
ZHANG Shunmiao; CHEN Minglong
School of Information Science and Engineering, Fujian University of Technology
关键词:
非侵入式负荷分解;  k-means聚类;  状态码;  注意力机制
Keywords:
non-intrusive load decomposition;  k-means clustering;  status code;  attention mechanism
分类号:
TM714
DOI:
10.3969/j.issn.1672-4348.2020.04.006
文献标志码:
A
摘要:
针对非侵入式负荷分解准确率低的问题，提出一种新的非侵入式负荷分解方法。首先，针对在训练时难以对含有大量数据的电器设备的工作状态做标签问题，设计了引入轮廓系数和平方误差和共同作为评价指标的k-means聚类来确定负荷状态数，构建了状态码表示所有电器的运行状态。其次，利用卷积层和双向长短期记忆网络对特征进行提取，并引入注意力机制选取对分解任务重要性程度高的电器状态码，然后，通过全连接层进行分类，得到各时刻下的状态码，进而得到各用电设备实际功率。最后利用公开AMPds2数据集进行验证，结果表明所提方法，具有较高的负荷分解准确率。
Abstract:
Aiming at the problem that the accuracy of non-intrusive load decomposition still needs to be improved, a non-intrusive load decomposition method was proposed. First, aiming at the difficulty in labelling the operation status of electric equipment with a large amount of data during training, k-means clustering introducing the silhouette coefficient and the sum of squared error as evaluation indicators was designed to determine the number of load status. A status code was constructed to represent the operation state of all electrical appliances. Second, features were extracted by means of ConvBiLSTM(convolution layer and bidirectional long short-term memory network). Attention mechanism was introduced to select the electrical status codes with high importance to the decomposition task, and then these codes were classified through the full connection layer to obtain the status codes at each moment, and then the actual power of each electrical equipment was obtained. Finally, the public AMPds2 data set was used to verify the proposed method. Results show that the proposed method has higher load decomposition accuracy.

参考文献/References:

[1] HART G. Nonintrusive appliance load monitoring[J]. Proceedings of the IEEE, 1992, 80(12): 1870-1891.[2] 阿蓉. 场景自适应的非侵入式负荷识别算法研究[D]. 武汉: 华中科技大学, 2018. [3] 曹敏, 魏龄, 邹京希, 等. 基于暂态过程的非侵入式负荷监测研究[J]. 水电能源科学, 2018, 36(8): 177-180.[4] 韩笑, 邓春宇, 张玉天, 等. 基于高频数据V-I特性的延时反馈负荷在线快速辨识算法[J]. 电力系统自动化, 2019, 43(9): 108-116.[5] 孙毅, 张璐, 赵洪磊, 等. 基于动态自适应粒子群算法的非侵入式家居负荷分解方法[J]. 电网技术, 2018, 42(6): 1819-1826.[6] LIU B, LUAN W, YU Y. Dynamic time warping based non-intrusive load transient identification[J]. Applied Energy, 2017, 195: 634-645.[7] MAKONIN S, POPOWICH F, BAJI〖KG-*4〗C〖DD(-*3〗′〖DD)〗 I, et al. Exploiting HMM sparsity to perform online real-time nonintrusive load monitoring[J]. IEEE Transactions on Smart Grid, 2016, 7(6): 2575-2585.[8] 徐青山, 娄藕蝶, 郑爱霞, 等. 基于近邻传播聚类和遗传优化的非侵入式负荷分解方法[J]. 电工技术学报, 2018, 33(16): 3868-3878.[9] 彭显刚, 郑凯, 林哲昊, 等. 基于谱图理论的居民用户非侵入式负荷分解[J]. 电网技术, 2018, 42(8): 2674-2680.[10] ZHANG C, ZHONG M, WANG Z, et al. Sequence-to-point learning with neural networks for non-intrusive load monitoring. [C]∥2018 Thirty-Second AAAI Conference on Artificial Intelligence, 2018.[11] 刘耀先, 孙毅, 李彬, 等. 基于边缘嵌入深度学习的非侵入式负荷分解方法[J]. 电网技术, 2019, 43(12): 4329-4337.[12] LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.[13] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.[14] SCHUSTER M, PALIWAL K. Bidirectional recurrent neural networks[J]. IEEE Transactions on Signal Processing, 1997, 45(11): 2673-2681.[15] MAKONIN S, POPOWICH F, BARTRAM L, et al. AMPds: a public dataset for load disaggregation and eco-feedback research[C]∥2013 IEEE Electrical Power & Energy Conference. Halifax, NS, Canada: IEEE, 2013: 1-6. [16] 任文龙, 许刚. 基于深度序列翻译模型的非侵入式负荷分解方法[J]. 电网技术, 2020, 44(1): 27-37.[17] BATRA N, KELLY J, PARSON O, et al. NILMTK: an open source toolkit for non-intrusive load monitoring[EB/OL]. 2014: arXiv:1404.3878[stat.AP]. https:∥arxiv.org/abs/1404.3878.[18] KELLY J, KNOTTENBELT W. Neural NILM: deep neural networks applied to energy disaggregation[C]∥Proceedings of the 2nd ACM International Conference on Embedded Systems for Energy-Efficient Built Environments-BuildSys 15. New York: ACM Press, 2015: 55-64.

更新日期/Last Update: 2020-08-25