一种非侵入式低功耗高精度的用电器识别算法

doi:10.19665/j.issn1001-2400.2023.01.017

摘要/Abstract

摘要：

非侵入式用电器识别是实现泛在电力物联网客户侧智能感知的关键技术。针对现有非侵入式用电器识别系统计算复杂度过高,不利于产业化的问题,提出了一种基于长短期记忆网络的低成本、低功耗、高精度且可应用于嵌入式微控制器的非侵入式用电器识别算法和系统。首先,对总线上的电流进行同步采集;其次,利用所提出的多参数的检测方法判断出用电器的投切事件;再次,利用长短期记忆网络,对投切时间点前后变化的数据进行识别,得到投切事件的用电器种类;最后,结合累积和判断出当前的用电器的种类和数量。仿真和实测结果表明,只需要针对单个用电器进行少量数据的训练,即可在嵌入式微控制器系统中对组合用电器实现高达99.6%的平均识别准确度,且系统功耗小于1.5 W。这种算法可进一步应用于统计各个用电器的投切时间点、使用时长和功耗总和等工作情况,为智能电网提供精细化的用户用电信息,并为电网能源管理和优化提供重要参考。

关键词: 物联网, 长短期记忆网络, 识别, 电参数测量

Abstract:

Non-intrusive electrical appliance identification is the key technology to realize customer-side intelligent sensing in the ubiquitous power Internet of Things.Aiming at the problem that the calculation complexity of the existing non-intrusive electrical appliance identification system is too high and is not conducive to industrialization,a low-cost,low-power consumption,high-precision non-intrusive electrical appliance identification system and the algorithm based on the Long Short-Term Memory(LSTM) network are proposed,and they can be applied to an embedded microcontroller.First,the current on the bus is collected synchronously.Second,the proposed multi-parameter detection method is used to judge the switching event of the electrical appliance.Third,the LSTM network is used to process the data before and after the switching time point,and the type of electrical appliance which is switching is obtained.Finally,the current types and quantities of electrical appliances are judged by the cumulative sum.Simulation and measurement results show that only a small amount of data training for a single electric appliance is needed,and that the recognition accuracy of combined electric appliances can be up to 99.6% in the proposed embedded microcontroller system,in which the power consumption is less than 1.5 watts.The proposed algorithm can be further applied to the statistics of the switching time point,service time and total power consumption of each electrical appliance,which provides refined user power consumption information for the smart grid.and provides an important reference for the energy management and optimization in the smart grid.

Key words: internet of things, LSTM, identification, electrical parameter measurement

中图分类号:

TM714

吴拨云,顾文杰,何先灯. 一种非侵入式低功耗高精度的用电器识别算法[J]. 西安电子科技大学学报, 2023, 50(1): 149-157.

WU Boyun,GU Wenjie,HE Xiandeng. Low-power consumption high-precision non-intrusive electrical appliance identification algorithm[J]. Journal of Xidian University, 2023, 50(1): 149-157.

图/表 12

图1

图2

图3

图4

图5

表1

表2

图6

图7

表3

表4

表5

参考文献 19

[1]	ALSALEMI A, RAMADAN M, BENSAALI F, et al. Endorsing Domestic Energy Saving Behavior Using Micro-Moment Classification[J]. Applied Energy, 2019, 250:1302-1311. doi: 10.1016/j.apenergy.2019.05.089
[2]	蒋忠远, 张秀艳, 马子玥. 智能电网中电能存储系统的容量优化[J]. 西安电子科技大学学报, 2015, 42(6):106-112.
	JIANG Zhongyuan, ZHANG Xiuyan, MA Ziyue. Capacity Optimization for Power Storage Systems in Smart Grids[J]. Journal of Xidian University, 2015, 42(6):106-112.
[3]	刘迪迪, 孙浩天, 肖佳文, 等. 智能电网中终端用户的双向能量交易算法[J]. 西安电子科技大学学报, 2021, 48(3):131-137.
	LIU Didi, SUN Haotian, XIAO Jiawen, et al. Two-Way Energy Trading Algorithms for the End-User in the Smart Grid.Journal of Xidian University[J]. 2021, 48(3):131-137.
[4]	CZÉTÁNY L, VÁMOS V, HORVÁTH M, et al. Development of Electricity Consumption Profiles of Residential Buildings Based on Smart Meter Data Clustering[J]. Energy and Buildings, 2021, 252:111376. doi: 10.1016/j.enbuild.2021.111376
[5]	ZHOU Y, LI F, LIU L, et al. Non-Intrusive Load Monitoring Method Based on the Time-Segmented State Probability[J]. Energy Reports, 2022, 8(S4):1418-1423. doi: 10.1016/j.egyr.2022.02.021
[6]	PARASKEVAS I, BARBAROSOU M, FITTON R, et al. Domestic Smart Metering Infrastructure and A Method for Home Appliances Identification Using Low-Rate Power Consumption Data[J]. IET Smart Cities, 2021, 3(2):91-106. doi: 10.1049/smc2.v3.2
[7]	FANG Z, ZHAO D, CHEN C, et al. Non-Intrusive Appliance Identification with Appliance-Specific Networks[C]// 2019 IEEE Industry Applications Society Annual Meeting.Piscataway:IEEE, 2019:1-8.
[8]	YAN D, JIN Y, SUN H, et al. Household Appliance Recognition through A Bayes Classification Model[J]. Sustainable Cities and Society, 2019, 46:101393. doi: 10.1016/j.scs.2018.12.021
[9]	HIMEUR Y, ALSALEMI A, BENSAALI F, et al. Smart Non-Intrusive Appliance Identification Using A Novel Local Power Histogramming Descriptor with An Improved K-nearest Neighbors Classifier[J]. Sustainable Cities and Society, 2021, 67:102764. doi: 10.1016/j.scs.2021.102764
[10]	周晓, 李永清, 张有兵. 基于ELM的非侵入式电力负荷识别算法[J]. 高技术通讯, 2020, 30(10):1018-1024.
	ZHOU Xiao, LI Yongqing, ZHANG Youbing. Identification Algorithm for Non-Intrusive Power Load Based on ELM[J]. Chinese High Technology Letters, 2020, 30(10):1018-1024.
[11]	UTOMO D, HSIUNG P A. Anomaly Detection at the IoT Edge Using Deep Learning[C]// 2019 IEEE International Conference on Consumer Electronics - Taiwan(ICCE-TW).Piscataway:IEEE, 2019:1-2.
[12]	周明, 宋旭帆, 涂京, 等. 基于非侵入式负荷监测的居民用电行为分析[J]. 电网技术, 2018, 42(10):3268-3274.
	ZHOU Ming, SONG Xufan, TU Jing, et al. Residential Electricity Consumption Behavior Analysis Based on Non-Intrusive Load Monitoring[J]. Power System Technology, 2018, 42(10):3268-3274.
[13]	汪颖, 杨维, 肖先勇, 等. 基于U-I轨迹曲线精细化识别的非侵入式居民负荷监测方法[J]. 电网技术, 2021, 45(10):4104-4113.
	WANG Ying, YANG Wei, XIAO Xianyong, et al. Non-Intrusive Residential Load Monitoring Method Based on Refined Identification of U-I Trajectory Curve[J]. Power System Technology, 2021, 45(10):4104-4113.
[14]	曲朝阳, 于华涛, 郭晓利. 基于开启瞬时负荷特征的家电负荷识别[J]. 电工技术学报, 2015, 30(S1):358-364.
	QU Zhaoyang, YU Huatao, GUO Xiaoli. The Recognition of Appliances Instantaneous Load[J]. Transactions of China Electrotechnical Society, 2015, 30(S1):358-364.
[15]	祁兵, 程媛, 武昕. 基于Fisher有监督判别的非侵入式居民负荷辨识方法[J]. 电网技术, 2016, 40(8):2484-2490.
	QI Bing, CHENG Yuan, WU Xin. Non-Intrusive Household Appliance Load Identification Method Based on Fisher Supervised Discriminant[J]. Power System Technology, 2016, 40(8):2484-2490.
[16]	NWAKANMAC I, ISLAM F B, MAHARANI M P, et al. IoT-Based Vibration Sensor Data Collection and Emergency Detection Classification using Long Short Term Memory(LSTM)[C]// 2021 International Conference on Artificial Intelligence in Information and Communication(ICAIIC).Piscataway:IEEE, 2021:273-278.
[17]	ZHAO S, WANG C, BIAN X. Research on Harmonic Detection Based on Wavelet Threshold and FFT Algorithm[J]. Systems Science & Control Engineering An Open Access Journal, 2018, 6(3):339-345.
[18]	ANDERSON K, OCNEANU A, BENITEZ D, et al. BLUED:A Fully Labeled Public Dataset for Event-based Nonintrusive Load Monitoring Research[C]// Proceedings of the 2nd KDD Workshop on Data Mining Applications in Sustainability(SustKDD). New York: ACM, 2012:1-5.
[19]	SINKEVYCH O, BOYKO Y, RECHYNSKYI O, et al. Embedding Sequence Model in STM32 Based Neuro-Controller[C]// 2021 IEEE 12th International Conference on Electronics and Information Technologies(ELIT).Piscataway:IEEE, 2021:113-118.

编号	用电器类型	A_cc/%	P_re/%	R_ec/%	F₁
1	冰箱	100.00	100.00	100.00	1.000
2	后院灯	100.00	100.00	100.00	1.000
3	吹风机	100.00	100.00	100.00	1.000
4	浴室顶灯	98.40	98.90	97.80	0.983
5	卧室灯	97.30	94.20	100.00	0.971
6	厨房切碎机	99.80	100.00	95.20	0.976
平均		99.25	98.85	98.83	0.988

编号	用电器类型	ND-LSTM	Adaboost-BP	UIF-NILM	Fisher判别
1	冰箱	100.00	100.00	100.00	71.40
2	后院灯	100.00	100.00	98.10	100.00
3	吹风机	100.00	100.00	100.00	100.00
4	浴室顶灯	98.40	96.90	96.10	71.90
5	卧室灯	97.30	83.30	100.00	83.30
6	厨房切碎机	99.80	80.00	100.00	100.00
平均识别率		99.25	93.37	99.03	87.77

编号	用电器名称	cos(?(1))	D(3)	D(5)	P_max/W	P_av/W	I/mA	V_t
1	1号自制用电器	0.20	0.31	0.14	1.46	0.98	43	0.55
2	LED灯	1.00	0.15	0.09	1.48	0.99	45	1.14
3	2号自制用电器	0.80	0.28	0.15	1.45	0.97	42	0.93
4	电风扇	0.50	0.20	0.12	1.49	0.99	202	2.26
5	47 kΩ电阻	1.00	0.00	0.00	1.48	0.98	5	1.00
6	电吹风	0.90	0.08	0.01	1.45	0.96	8 550	89.68
7	节能灯	0.27	0.30	0.08	1.47	0.97	49	0.71
8	电源	-0.30	0.40	0.10	1.48	0.99	18	8.92

编号	用电器名称	识别准确度
编号	用电器名称	ND-LSTM	UIF-NILM
1	1号自制用电器	99.4	96.2
2	LED灯	99.7	96.7
3	2号自制用电器	98.9	98.1
4	电风扇	99.8	99.2
5	47 kΩ电阻	99.9	90.5
6	电吹风	100.0	96.2
7	节能灯	99.7	97.3
8	电源	99.4	98.1
	平均	99.6	96.5

编号	用电器名称	系统功耗
编号	用电器名称	ND-LSTM	UIF-NILM
1	1号自制用电器	1.47	5.9
2	LED灯	1.47	5.8
3	2号自制用电器	1.46	6.1
4	电风扇	1.49	6.2
5	47 kΩ电阻	1.48	5.9
6	电吹风	1.46	6.0
7	节能灯	1.47	6.1
8	电源	1.46	5.8
	平均	1.47	5.97