Multi-Working Condition Energy Consumption Anomaly Detection Method for Office Building Lighting Sockets Based on LSTM-AE

doi:10.12141/j.issn.1000-565X.240575

Abstract

Abstract:

Anomaly detection of energy consumption in building lighting and socket systems can effectively improve energy efficiency. It holds significant importance for the implementation of building energy optimization measures and the realization of energy-saving management and control. Since the energy consumption of building lighting and plug load systems is heavily influenced by the random behavior of building occupants, and given the challenges posed by noisy time-series data and difficulty in feature extraction, this study proposed an unsupervised anomaly detection method that integrates operating condition classification with deep learning, aiming to enhance the accuracy and robustness of energy consumption anomaly identification. First, the decision tree algorithm was employed to classify the energy data based on attributes such as working days vs. non-working days and working hours vs. non-working hours. Then, for each identified condition, a long short-term memory autoencoder (LSTM-AE) model was constructed to detect anomalies. This model learns to reconstruct normal data and calculates the reconstruction error. By setting differentiated thresholds, it enables energy consumption anomaly detection under unlabeled data conditions. Using 578 days of hourly lighting and socket energy consumption data from an office building located in a hot-summer and warm-winter region, the study conducted model training and hyperparameter optimization experiments. Results indicate that the number of iterations, the number of neurons, and the activation function have significant effects on the model’s performance. Energy data during working days demonstrate greater stability than those on non-working days, resulting in higher detection accuracy. The proposed method achieves average precision, recall, and F₁ of 91.23%, 90.87%, and 90.80%, respectively, across four typical operating conditions, demonstrating its effectiveness in detecting energy anomalies in building lighting and socket systems.

Key words: energy consumption anomaly detection, deep learning, lighting socket system, reconstruction error, unsupervised learning

CLC Number:

TU831

CHEN Cheng, WANG Miao, WANG Xinyao, GAO Zhiming, ZHOU Xuan, YAN Junwei. Multi-Working Condition Energy Consumption Anomaly Detection Method for Office Building Lighting Sockets Based on LSTM-AE[J]. Journal of South China University of Technology(Natural Science Edition), 2025, 53(9): 117-126.

Figures/Tables 16

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

Fig.8

Fig.9

Fig.10

Fig.11

Fig.12

Fig.13

Table 3

Table 4

Fig.14

References 23

[1]	LUO S， HU W， LIU W，et al ．Study on the decarbonization in China's power sector under the background of carbon neutrality by 2060 ［J］．Renewable and Sustainable Energy Reviews，2022， 166：112618/1-18.
[2]	WANG M， XU D， BAI Y，et al ．Dynamic investigation on potassium migration and transformation during biochar combustion and its correlation with combustion reactivity［J］．Fuel，2023，340：127540/1-11.
[3]	中国建筑节能协会．2022中国建筑能耗与碳排放研究报告［R］．北京：中国建筑节能协会，2022.
[4]	AHMAD A S， HASSAN M Y， ABDULLAH M P，et al ．A review on applications of ANN and SVM for building electrical energy consumption forecasting ［J］．Renewable and Sustainable Energy Reviews，2014，33：102-109.
[5]	AMASYALI K， ELGOHARY N ．Building lighting energy consumption prediction for supporting energy data analytics ［J］．Procedia Engineering，2016，145：511-517.
[6]	李海涛，王瑞敏，董卫宇，等．一种基于GRU的半监督网络流量异常检测方法［J］．计算机科学，2023，50（3）：380-390.
	LI Haitao， WANG Ruimin， DONG Weiyu，et al ．Semi-supervised network traffic anomaly detection method based on GRU ［J］．Computer Science，2023，50（3）：380-390.
[7]	SERRANO-GUERRERO X， ESCRIVÁ-ESCRIVÁ G， LUNA-ROMERO S，et al ．A time-series treatment method to obtain electrical consumption patterns for anomalies detection improvement in electrical consumption profiles ［J］．Energies， 2020，13（5）：1046/1-23.
[8]	SERRANO-GUERRERO X， ESCRIVÁ-ESCRIVÁ G， ROLDÁN-BLAY C ．Statistical methodology to assess changes in the electrical consumption profile of buildings ［J］．Energy and Buildings，2018，164：99-108.
[9]	ZHANG L， WAN L， XIAO Y，et al ．Anomaly detection method of smart meters data based on GMM-LDA clustering feature learning and PSO support vector machine［C］∥ Proceedings of 2019 IEEE Sustainable Power and Energy Conference．Beijing：IEEE，2019：2407-2412.
[10]	PUNMIYA R， CHOE S ．Energy theft detection using gradient boosting theft detector with feature engineering-based preprocessing ［J］．IEEE Transactions on Smart Grid，2019，10（2）： 2326-2329.
[11]	HOLLINGSWORTH K， ROUSE K， CHO J，et al ．Energy anomaly detection with forecasting and deep learning ［C］∥ Proceedings of 2018 IEEE International Conference on Big Data．Seattle：IEEE，2018：4921-4925.
[12]	AHMED N， SAHA A K， NOMAN M A AL，et al ．Deep learning-based natural language processing in human-agent interaction：applications，advancements and challenges［J］．Natural Language Processing Journal，2024，9：100112/1-25.
[13]	FENZA G， GALLO M， LOIA V ．Drift-aware metho-dology for anomaly detection in smart grid ［J］．IEEE Access，2019，7：9645-9657.
[14]	SILVA A DA， GUARANY I S， ARRUDA B，et al ．A method for anomaly prediction in power consumption using long short-term memory and negative selection ［C］∥ Proceedings of 2019 IEEE International Symposium on Circuits and Systems．Sapporo：IEEE，2019：1-5.
[15]	WANG X， ZHAO T， LIU H，et al ．Power consumption predicting and anomaly detection based on long short-term memory neural network ［C］∥ Proceedings of 2019 IEEE the 4th International Conference on Cloud Computing and Big Data Analysis．Chengdu：IEEE，2019：487-491.
[16]	XU C， CHEN H ．Abnormal energy consumption detection for GSHP system based on ensemble deep learning and statistical modeling method ［J］．International Journal of Refrigeration，2020，114：106-117.
[17]	ZHENG Z， YANG Y， NIU X，et al ．Wide and deep convolutional neural networks for electricity-theft detection to secure smart grids ［J］．IEEE Transactions on Industrial Informatics，2018，14（4）：1606-1615.
[18]	LI S， HAN Y， YAO X，et al ．Electricity theft detection in power grids with deep learning and random fo-rests ［J］．Journal of Electrical and Computer Enginee-ring，2019，2019：4136874/1-12.
[19]	PRIYADARSINI M， SONEKAR N ．A CNN-based approach for anomaly detection in smart grid systems ［J］．Electric Power Systems Research，2025，238：111077/1-10.
[20]	PYDI D P， ADVAITH S ．Attention boosted autoencoder for building energy anomaly detection ［J］．Energy and AI，2023，14： 100292/1-11.
[21]	MASCALI L， SCHIERA D S， EIRAUDO S，et al ．A machine learning-based anomaly detection framework for building electricity consumption data ［J］．Sustainable Energy Grids and Networks，2023，36：101194/1-14.
[22]	RUFF L， KAUFFMANN J R， VANDERMEULEN R A，et al ．A unifying review of deep and shallow anomaly detection ［J］．Proceedings of the IEEE，2021，109（5）：756-795.
[23]	周欣，燕达，任晓欣，等．大型办公建筑照明能耗实测数据分析及模型初探［J］．照明工程学报，2013，24（4）：14-23.
	ZHOU Xin， YAN Da， REN Xiaoxin，et al ．Data analysis and modeling of lighting energy use in large office buildings ［J］．China Illuminating Engineering Journal，2013，24（4）：14-23.

工况	数据量		异常数据量		总数据量	异常率/%
工况	训练集	验证集	训练集	验证集	总数据量	异常率/%
1	2 701	1 159	102	3	3 860	2.72
2	3 781	1 623	34	5	5 404	0.72
3	1 245	535	17	0	1 780	0.95
4	1 745	747	26	3	2 492	1.08

工况	异常点数量	误差阈值	P/%		R/%		F₁/%
工况	异常点数量	误差阈值	AE	LSTM-AE	AE	LSTM-AE	AE	LSTM-AE
1	95	579.47	91.17	90.47	97.89	100.00	94.41	95.00
2	40	207.61	86.04	92.31	92.50	90.00	89.15	91.14
3	20	196.64	78.26	100.00	90.00	85.00	83.72	91.89
4	28	167.33	78.13	82.14	89.28	88.46	83.33	85.18