[1]刘 辉,江煦成. 基于多任务学习的非侵入式负荷分解[J].湖北工业大学学报,2023,(2):1-6.
　LIU Hui,JIANG Xucheng. Non-intrusive Load Decomposition Based on Multi-task Learning[J].,2023,(2):1-6.
点击复制

 基于多任务学习的非侵入式负荷分解()
分享到：

《湖北工业大学学报》[ISSN:1003-4684/CN:42-1752/Z]

卷:
期数:
2023年第2期
页码:
1-6
栏目:
湖北工业大学学报
出版日期:
2023-04-30

文章信息/Info

Title:
 Non-intrusive Load Decomposition Based on Multi-task Learning
文章编号:
1003-4684(2023)02-0001-06
作者:
 刘 辉;  江煦成
 湖北工业大学太阳能高效利用湖北省协同创新中心， 湖北 武汉 430068
Author(s):
 LIU Hui;  JIANG Xucheng
 Hubei Collaborative Innovation Centerfor Efficient Use of Solar Energy,Hubei Univ. of Tech., Wuhan 430068,China
关键词:
 ; 非侵入式负荷分解; 多任务学习; MMoE; seq2point
Keywords:
 ; non-intrusive load decomposition;  multi-task learning;  mmoe;  seq2point
分类号:
TM714
文献标志码:
A
摘要:
 为解决目前非侵入式负荷分解研究中存在的模型数量多及训练时间长等问题，将多任务学习引入到非侵入式负荷分解研究中，提出一种基于多门多专家模型的非侵入式负荷分解方法，首先通过seq2point模型将用电设备的功率分解转换为总功率序列与用电设备在序列中点时刻功率值的映射关系，其次利用MMoE模型的门控函数及共用的Expert网络组兼顾不同用电设备功率分解任务的独特性和关联性，最终通过单个MMoE模型同时完成多个用电设备的功率分解。在公开数据集上进行验证，测试算例验证了方法的有效性。
Abstract:
 Non-intrusive load decomposition is one of the key technologies for obtaining user electricity information in the construction of the power grid. In order to solve the problems of the large number of models and long training time in the current non-intrusive load decomposition research, this paper introduces multi-task learning In the study of non-intrusive load decomposition, a non-intrusive load decomposition method based on the Multi-gate Mixture of-Experts (MMoE) model is proposed. First, the power of the electrical equipment is decomposed and converted through the seq2point model. It is the mapping relationship between the total power sequence and the power value of the electrical equipment at the midpoint of the sequence. Secondly, the gating function of the MMoE model and the shared Expert network group are used to take into account the uniqueness and relevance of the power decomposition tasks of different electrical equipment, and finally pass A single MMoE model completes the power decomposition of multiple electrical equipment at the same time. This article verifies the method proposed in this article on the public data set, and the test case verifies the effectiveness of the method in this article.

参考文献/References:

[1] KELLY J, KNOTTENBELT W J. Neural NILM:deep neural networksapplied to energy disaggregation[C].∥Proceedings of the 2nd ACMInternational Conference on Embedded Systems for Energy-Efficient Built Environments .Seoul,Korea: ACM，2015， 55-64.
[2] 刘仲民,侯坤福,高敬更.基于时间卷积神经网络的非侵入式居民用电负荷分解方法[J].电力建设,2021,42(03):97-106. 
[3] 徐晓会,赵书涛,崔克彬.基于卷积块注意力模型的非侵入式负荷分解算法[J/OL].电网技术:1-8[2021-09-06].https://doi.org/10.13335/j.1000-3673.pst.2020.1487.].
[4] 王轲，钟海旺，余南鹏，等．基于 seq2seq 和 Attention 机制的居民用户非侵入式负荷分解[J]．中国电机工程学报，2019，39(01)：75-83．
[5] YANG Y, HOSPEDALES T. Deep multi-task representation learning: a tensor factorisation approach[J/OL]. （2022-03-09）.https:∥arxiv.org/abs/1605.06391.
[6] MISRA I, SHRIVASTAVA A, GUPTA A, et al. Cross-stitch networks for multi-task learning[C].∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2016.
[7] MA J, ZHAO Z, YI X, et al. Modeling task relationships in multi-task learning with multi-gate mixture-of-experts[C]. ACM, 2018. [8] Rich Caruana. 1998. Multitask learning. In Learning to learn. Springer, 95-133.
[8] CARUANA R. Multitask learning: a knowledge-based source of inductive bias[C].∥Machine Learning, Proceedings of the Tenth International Conference, University of Massachusetts, Amherst, MA, USA, June 27-29, 1993. DBLP, 1993.
[9] EIGEN D, RANZATO M, SUTSKEVER I. Learning factored representations in a deep mixture of experts[J/OL].（2022-03-09）.https://arxiv.org/abs/1312.4314.
[10] JACOBS R, JORDAN M, NOWLAN S, et al. Adaptive mixtures of local experts[J]. Neural Computation, 2014, 3(01):79-87.
[11] KOLTER J Z, Johnson M J. REDD: A public data set for energydisaggregation research[C].∥Workshop on Data Mining Applicationsin Sustainability(SIGKDD), San Diego, CA. 2011, 25(Citeseer): 59-62.

相似文献/References:

[1]熊韧,曹海印,王焱清,等.非牛顿润滑静压轴承的节流器流量方程修正[J].湖北工业大学学报,2019,34(5):6.
　XIONG Ren,CAO Haiyin,WANG Yanqing,et al.Modified restrictor flow equations of hydrostatic bearings ubricated by non-Newtonian fluids[J].,2019,34(2):6.
[2]王照远,曹 民,王 毅,等. 场景与数据双驱动的隧道图像拼接方法[J].湖北工业大学学报,2020,(4):11.
　WANG Zhaoyuan,CAO Min,WANG Yi,et al. Tunnel Image Stitching Method based on Scene and Data[J].,2020,(2):11.
[3]潘 健,梁佳成,陈凤娇,等. 单电流闭环多重PR控制的LCL型逆变器[J].湖北工业大学学报,2020,(4):16.
　PAN Jian,LIANG Jiacheng,CHEN Fengjiao,et al. Design of LCL Grid Connected Inverter based on Single Closed Loop Control and Multiple PR Controllers[J].,2020,(2):16.
[4]王晓光,赵 萌,文益雪,等. 定子闭口槽结构对永磁电机齿槽转矩影响分析[J].湖北工业大学学报,2020,(4):25.
　WANG Xiaoguang,ZHAO Meng,WEN Yixue,et al. Study on Cogging Torque and Vibration Noise of Permanent Magnet Motor with Segmental Stator and Closed-Slot[J].,2020,(2):25.
[5]宇 卫,凃玲英,陈 健. 风电场集中接入对集电线电流保护的影响[J].湖北工业大学学报,2020,(4):29.
　YU Wei,TU Lingying,CHEN Jian. Effect of the Collective Line Current Protection when Wind Farms are Centralized Accessed to the Power System[J].,2020,(2):29.
[6]廖政斌,王泽飞,祝 珊. 二惯量系统谐振在线抑制及相位补偿[J].湖北工业大学学报,2020,(4):34.
　LIAO Zhengbin,WANG Zefei,ZHU Shan. Online Resonance Suppression and Phase Compensation for Double Inertia System[J].,2020,(2):34.
[7]王 欣,游 颖,姜天翔,等. 面向3D打印过程的产品工艺设计和优化[J].湖北工业大学学报,2020,(4):39.
　WANG Xin,YOU Ying,JIANG Tianxiang,et al. Product Process Design and Optimization for 3D Printing Processes[J].,2020,(2):39.
[8]冉晶晶,文 红,罗雅梅,等. 全自动样品前处理平台及其控制系统[J].湖北工业大学学报,2020,(4):43.
　RAN Jingjing,WEN Hong,LUO Yamei,et al. Research on Automatic Sample Preprocessing Platform and its Control System[J].,2020,(2):43.
[9]杨 磊,马志艳,石 敏,等. 基于模糊PID的小型冷库过热度控制方法[J].湖北工业大学学报,2020,(4):43.
　YANG Lei,MA Zhiyan,SHI Min,et al. Research on Superheat Control Method of Small Cold Storage based on Fuzzy PID[J].,2020,(2):43.
[10]黄 晶,周细枝,周业望. 动态注塑成型模具的设计与实验研究[J].湖北工业大学学报,2020,(4):52.
　HUANG Jing,ZHOU Xizhi,ZHOU Yewang. Design and Experimental Study of Dynamic Injection Molding[J].,2020,(2):52.

备注/Memo

备注/Memo:
［收稿日期］ 2022-03-10
［基金项目］ 国家自然科学基金项目（61903129）
［第一作者］ 刘 辉（1962-），男，湖北武汉人，湖北工业大学教授，研究方向为电网控制与智能制造
［通信作者］ 江煦成（1996-），男，湖北黄冈人，湖北工业大学硕士研究生，研究方向为非侵入式负荷分解
更新日期/Last Update: 2023-04-26