Non-intrusive load identification for public buildings based on MSCNN-BiGRU-MLP model

doi:10.3969/j.issn.2097-0706.2025.03.003

Abstract

Abstract:

In the energy management of public buildings， load identification plays a critical role in optimizing energy utilization and reducing energy consumption. Traditional load monitoring methods are primarily intrusive， relying on hardware equipment or macro-level load characteristics， which fail to meet the refined management requirements of modern intelligent buildings and smart cities. To address the challenges posed by the diversity and uncertainty of public building loads， a non-intrusive load identification method was proposed based on multi-scale convolutional neural network （MSCNN）， bidirectional gated recurrent unit（BiGRU）， and multilayer perceptron（MLP）. The model integrated voltage-current（V-I） trajectory features， power features， and harmonic features to achieve classification and identification of typical socket-based loads in public buildings. MSCNN was employed to extract V-I trajectory features， capturing stable and "fingerprint-like" characteristics of equipment during operation. BiGRU was utilized for time-series modeling of power and harmonic features， revealing the dynamic characteristics of load signals. MLP was then applied to classify the fused features. Experiments on various common public building loads validated the effectiveness of the proposed model. The results showed that the MSCNN-BiGRU-MLP model achieved a load identification accuracy of 0.917 1， accurately identifying load types and maintaining high robustness under dynamic feature changes and high-frequency noise conditions.

Key words: non-intrusive load identification, multi-scale convolutional neural network, bidirectional gated recurrent unit, multilayer perceptron, public buildings, voltage-current（V-I） trajectory features, energy management

CLC Number:

TK01

YANG Lijie, DENG Zhenyu, CHEN Zuoshuang, HUANG Chao, JIANG Meihui, ZHU Hongyu. Non-intrusive load identification for public buildings based on MSCNN-BiGRU-MLP model[J]. Integrated Intelligent Energy, 2025, 47(3): 23-31.

Figures/Tables 9

Fig.1

Fig.2

Fig.3

Fig. 4

Fig.5

Fig.6

Table 1

Table 2

Table 3

References 25

[1]	邹风华, 朱星阳, 殷俊平, 等. “双碳” 目标下建筑能源系统发展趋势分析[J]. 综合智慧能源, 2024, 46(8): 36-40. doi: 10.3969/j.issn.2097-0706.2024.08.005
	ZOU Fenghua, ZHU Xingyang, YIN Junping, et al. Development trend analysis on building energy systems under "dual carbon" target[J]. Integrated Intelligent Energy, 2024, 46(8): 36-40. doi: 10.3969/j.issn.2097-0706.2024.08.005
[2]	DING C, KE J, LEVINE M, et al. Potential of artificial intelligence in reducing energy and carbon emissions of commercial buildings at scale[J]. Nature Communications, 2024, 15(1): 5916. doi: 10.1038/s41467-024-50088-4 pmid: 39004671
[3]	RAFIQ H, MANANDHAR P, RODRIGUEZ-UBINAS E, et al. A review of current methods and challenges of advanced deep learning-based non-intrusive load monitoring (NILM) in residential context[J]. Energy and Buildings, 2024, 305: 113890.
[4]	ANGELIS G F, TIMPLALEXIS C, KRINIDIS S, et al. NILM applications: Literature review of learning approaches, recent developments and challenges[J]. Energy and Buildings, 2022, 261: 111951.
[5]	李亦非, 王芳, 张雅静, 等. 考虑时空耦合特性的非侵入式综合能源系统多能设备负荷辨识方法[J]. 科学技术与工程, 2024, 24(26): 11283-11293.
	LI Yifei, WANG Fang, ZHANG Yajing, et al. Non-invasive comprehensive energy system multi energy equipment load identification method considering spatiotemporal coupling characteristics[J]. Science Technology and Engineering, 2024, 24(26): 11283-11293.
[6]	龚玉婕. 基于非侵入式负荷分解的办公建筑用电行为研究[D]. 广州: 广东财经大学, 2022.
	GONG Yujie. Research on electricity consumption behavior of office buildings based on non-intrusive load decomposition[D]. Guangzhou: Guangdong University of Finance & Economics, 2022.
[7]	樊颜搏, 熊亚选, 李想, 等. 基于遗传算法的建筑用能多目标优化应用进展[J]. 综合智慧能源, 2024, 46(9): 69-85. doi: 10.3969/j.issn.2097-0706.2024.09.009
	FAN Yanbo, XIONG Yaxuan, LI Xiang, et al. Advancement in multi-objective optimization for building energy use based on genetic algorithms[J]. Integrated Intelligent Energy, 2024, 46(9): 69-85. doi: 10.3969/j.issn.2097-0706.2024.09.009
[8]	杨秀, 吴吉海, 孙改平, 等. 基于深度学习和迁移学习的公共楼宇非侵入式负荷分解[J]. 电网技术, 2022, 46(3): 1160-1169.
	YANG Xiu, WU Jihai, SUN Gaiping, et al. Non-intrusive load decomposition of public buildings based on deep learning and transfer learning[J]. Power System Technology, 2022, 46(3): 1160-1169.
[9]	孙健, 张云帆, 蔡潇龙, 等. 基于预测负荷的暖通空调系统优化调度[J]. 综合智慧能源, 2024, 46(3): 12-19. doi: 10.3969/j.issn.2097-0706.2024.03.002
	SUN Jian, ZHANG Yunfan, CAI Xiaolong, et al. Optimal scheduling of HVAC systems based on predicted loads[J]. Integrated Intelligent Energy, 2024, 46(3): 12-19. doi: 10.3969/j.issn.2097-0706.2024.03.002
[10]	NORFORD L K, LEEB S B. Non-intrusive electrical load monitoring in commercial buildings based on steady-state and transient load-detection algorithms[J]. Energy and Buildings, 1996, 24(1): 51-64.
[11]	AKBAR M K, AMAYRI M, BOUGUILA N. A novel non-intrusive load monitoring technique using semi-supervised deep learning framework for smart grid[J]. Building Simulation, 2024, 17(3): 441-457.
[12]	SILVA NOLASCO LDA, LAZZARETTI A E, MULINARI B M. DeepDFML-NILM: A new CNN-based architecture for detection,feature extraction and multi-label classification in NILM signals[J]. IEEE Sensors Journal, 2022, 22(1): 501-509.
[13]	HUANG L, CHEN S J, LING Z X, et al. Non-invasive load identification based on LSTM-BP neural network[J]. Energy Reports, 2021, 7: 485-492.
[14]	ZHOU X X, FENG J R, LI Y. Non-intrusive load decomposition based on CNN-LSTM hybrid deep learning model[J]. Energy Reports, 2021, 7: 5762-5771.
[15]	YUAN J, JIN R, WANG L D, et al. A Nonintrusive load identification method based on dualbranch attention GRU fusion network[J]. IEEE Transactions on Instrumentation and Measurement, 2024,73:1-10.
[16]	CHEN T, QIN H Y, LI X S, et al. A non-intrusive load monitoring method based on feature fusion and SE-ResNet[J]. Electronics, 2023, 12(8): 1909.
[17]	DJORDJEVIC S, SIMIC M. Nonintrusive identification and type recognition of household appliances based on the harmonic analysis of the steady-state current[J]. Electrical Engineering, 2023, 105(5): 3319-3328.
[18]	MEIER A, CAUTLEY D. Practical limits to the use of non-intrusive load monitoring in commercial buildings[J]. Energy and Buildings, 2021, 251: 111308.
[19]	裘星, 尹仕红, 张之涵, 等. 基于V-I轨迹与高次谐波特征的非侵入式负荷识别方法[J]. 电力工程技术, 2021, 40(6): 34-42.
	QIU Xing, YIN Shihong, ZHANG Zhihan, et al. Non-intrusive load identification method based on V-I trajectory and high-order harmonic feature[J]. Electric Power Engineering Technology, 2021, 40(6): 34-42.
[20]	石鑫, 刘奇央, 高峰. 深度神经网络在新型能源系统中的应用及展望[J]. 综合智慧能源, 2025, 47(2): 88-101. doi: 10.3969/j.issn.2097-0706.2025.02.009
	SHI Xin, LIU Qiyang, GAO Feng. Application and prospects of deep neural network in new energy systems[J]. Integrated Intelligent Energy, 2025, 47(2): 88-101. doi: 10.3969/j.issn.2097-0706.2025.02.009
[21]	ZAERI N, ASHOURI A, GUNAY H B, et al. Disaggregation of electricity and heating consumption in commercial buildings with building automation system data[J]. Energy and Buildings, 2022, 258: 111791.
[22]	刘航, 刘春阳, 赵浩然, 等. 基于深度学习的公共楼宇电-气负荷非侵入式分解[J]. 电网技术, 2023, 47(3): 1188-1197.
	LIU Hang, LIU Chunyang, ZHAO Haoran, et al. Non-intrusive disaggregation of electricity and gas load in public buildings based on deep learning[J]. Power System Technology, 2023, 47(3): 1188-1197.
[23]	刘一宁, 陈柏安, 杜鹏程, 等. 基于MDLOF-iForest和M-KNN-Slope的公共建筑负荷异常数据识别与修复[J]. 综合智慧能源:1-11(2025-03-05)[2025-03-10]. http://kns.cnki.net/kcms/detail/41.1461.tk.20250303.0944.002.html.
	LIU Yining, CHEN Baian, DU Pengcheng, et al. Detection and repair of abnormal load data of public buildings based on MDLOF-iForest and M-KNN-Slope[J]. Integrated Intelligent Energy,1-11(2025-03-05)[2025-03-10]. http://kns.cnki.net/kcms/detail/41.1461.tk.20250303.0944.002.html.
[24]	MULINARI B M, LINHARES R R, DE CAMPOS D P, et al. A new set of steady-state and transient features for power signature analysis based on V-I trajectory[C]//Proceedings of 2019 IEEE PES Innovative Smart Grid Technologies Conference—Latin America (ISGT Latin America),IEEE, 2019: 1-6.
[25]	DUC M L, BILIK P, MARTINEK R. Harmonics signal feature extraction techniques: A review[J]. Mathematics, 2023, 11(8): 1877.

变量	A	P	F₁
M₁	0.590 9	0.612 5	0.607 4
M₂	0.831 4	0.852 1	0.837 9
M₃	0.891 2	0.919 0	0.871 6
M₄	0.874 1	0.871 0	0.897 2
M₅	0.890 5	0.930 0	0.953 2
M₆	0.917 1	0.918 0	0.917 4

模型	A	P	F₁
SVM	0.621 4	0.603 2	0.596 3
Seq2Point	0.856 7	0.841 5	0.835 8
本文模型	0.917 1	0.918 0	0.917 4

负荷种类	A	P	R
空调	0.936 0	0.978 0	0.957 0
洗衣机	0.896 0	0.956 0	0.935 0
微波炉	0.979 0	0.979 0	1.000 0
笔记本电脑	0.978 0	0.978 0	0.957 0
白炽灯	0.784 0	0.889 0	0.870 0
平均值	0.914 0	0.956 0	0.944 0