Anomaly Detection of Complex Building Energy Consumption System Based on Machine Learning

doi:10.12141/j.issn.1000-565X.210422

Abstract

Abstract:

The research on the anomaly detection of the operation state of complex energy consumption system is of great significance to the safe, stable, efficient and energy-saving operation of buildings. However, due to the influence of many factors, such as the variety of external environmental factors of buildings, the uncertainty of insider behavior, the complexity of equipment operation data and so on, the detection of abnormal operation state often meets some difficulties in information feature extraction, abnormal operation state definition and so on. Based on the classification of operation parameters and by using the quantitative description of uncertainty by information entropy and the adaptability of unsupervised learning, this paper proposed a secondary clustering anomaly detection method for the operation state of complex energy consumption system based on information entropy and by integrating K-means and self-organizing mapping model. According to the actual monitoring data of the central air-conditioning system of a large office building in the hot summer and warm winter area, the external parameters were clustered for the first time through K-means, and the steps such as dividing the primary working conditions, calculating the information entropy of the operating parameters of the subsystem under each working condition, and determining the abnormal operating state by secondary clustering of self-organizing mapping model were used to realize the abnormal detection of the operating state of the complex energy consumption system of the building. The results show that the average intra class anomaly detection accuracy of the method proposed in this study is 97.45% under the operation condition of single machine and 96.70% under the operation condition of two machines. In addition, this study further discussed and analyzed the relationship between the number of classes and the relevant indicators of in class anomaly rate under different working conditions. It concludes that the in class anomaly detection rate of single and dual hosts increases with the decrease of the total number of operating states in their class. This study provides a new systematic idea and method for the abnormal operation state detection of complex energy consumption system under the background of building energy big data.

Key words: complex energy consumption system, information entropy, K-means, self-organizing maps model, anomaly detection

CLC Number:

TK01
TU831

ZHOU Xuan, WANG Xinyao, YAN Junwei, et al. Anomaly Detection of Complex Building Energy Consumption System Based on Machine Learning[J]. Journal of South China University of Technology(Natural Science Edition), 2022, 50(7): 144-154.

Figures/Tables 16

Table 1

Fig.1

Fig.2

Table 2

Fig.3

Table 3

Fig.4

Fig.5

Fig.6

Table 4

Table 5

Fig.7

Fig.8

Table 6

Fig.9

Fig.10

References 33

1	中国建筑节能协会．中国建筑能耗研究报告2020［J］．建筑节能（中英文），2021，49（2）：1-6.
	China Building Energy Conservation Association.China building energy consumption annual report 2020［J］．Journal of BEE，2021，49（2）：1-6.
2	XU C， CHEN H. A hybrid data mining approach for anomaly detection and evaluation in residential buildings energy data［J］．Energy and Buildings，2020，215：109864/1-13.
3	KATIPAMULA S， BRAMBLEY M R. Review article：methods for fault detection，diagnostics，and prognostics for building systems—a review［J］．HVAC&R Research，2005，11（1）：3-25.
4	苗露. 基于监测数据的建筑复杂用能设备运行性能预测与分析研究［D］．天津：天津大学，2014.
5	BUTTERWORTH H ．Handbook of energy efficiency in buildings［M］．Amsterdam：Elsevier，2019：597-673.
6	丁洪涛，刘海柱，殷帅. 我国公共建筑节能监管平台建设现状及趋势研究［J］．建设科技，2017（23）：10-11.
	DING Hongtao， LIU Haizhu， Yin Shuai. Research on the current situation and trend of public building energy conservation supervision platform construction in China［J］．Construction Science and Technology，2017（23）：10-11.
7	LEE K P， WU B H， PENG S L. Deep-learning-based fault detection and diagnosis of air-handling units［J］．Building and Environment，2019，157：24-33.
8	GUNAY B， SHEN W， YANG C. Characterization of a building’s operation using automation data：a review and case study［J］．Building and Environment，2017，118：196-210.
9	SUN L L， WU J H， JIA H Q，et al. Research on fault detection method for heat pump air conditioning system under cold weather［J］．Chinese Journal of Chemical Engineering，2017，25（12）：1812-1819.
10	VERBERT K， BABUSKA R， de SCHUTTER B. Combining knowledge and historical data for system-level fault diagnosis of HVAC systems［J］．Engineering Applications of Artificial Intelligence，2017，59：260-273.
11	LIANG Y X， MENG Q L， CHANG S. Fault diagnosis and energy consumption analysis for variable air volume air conditioning system： a case study［J］．Procedia Engineering，2017，205：834-841.
12	LI Y F， O'NEIL Z. An innovative fault impact analysis framework for enhancing building operations［J］．Energy and Buildings，2019，199：311-331.
13	LIU J H， LIU J Y， CHEN H X，et al. Abnormal energy identification of variable refrigerant flow air-conditioning systems based on data mining techniques［J］．Applied Thermal Engineering，2019，150：398-411.
14	GUO Y B， TAN Z H， CHEN H X，et al. Deep learning-based fault diagnosis of variable refrigerant flow air-conditioning system for building energy saving［J］．Applied Energy，2018，225：732-745.
15	PAN Y， ZHANG L M ．Data-driven estimation of building energy consumption with multi-source heterogeneous data［J］．Applied Energy，2020，268：114965/1-15.
16	PEÑA M， BISCARRI F， GUERRERO J I，et al. Rule-based system to detect energy efficiency anomalies in smart buildings，a data mining approach［J］．Expert Systems with Applications，2016， 56： 242-255.
17	FAN C， XIAO F， ZHAO Y，et al. Analytical investigation of autoencoder-based methods for unsupervised anomaly detection in building energy data［J］．Applied Energy，2018，211：1123-1135.
18	LUO X J， FONG K F， SUN Y J. Development of clustering-based sensor fault detection and diagnosis strategy for chilled water system［J］．Energy and Buildings，2019，186： 17-36.
19	LIU X， DING Y， TANG H，et al. A data mining-based framework for the identification of daily electricity usage patterns and anomaly detection in building electricity consumption data［J］．Energy and Buildings，2021，231：110601/1-22.
20	DU Z M， FAN B， JIN X Q，et al. Fault detection and diagnosis for buildings and HVAC systems using combined neural networks and subtractive clustering analysis［J］．Building and Environment，2014，73： 1-11.
21	SHI Z X， OBRIEN W. Development and implementation of automated fault detection and diagnostics for building systems：a review［J］．Automation in Construction，2019，104：215-229.
22	FAHIM M， FRAZ K， SILLITTI A. TSI：Time series to imaging based model for detecting anomalous energy consumption in smart buildings［J］．Information Sciences，2020，523：1-13.
23	BARUM H， BARRETT J， CLARK L O，et al. Entropy and information causality in general probabilistic theories［J］．New Journal of Physics，2010（3） 1-32.
24	AI Y T， GUAN J Y， FEI C W，et al. Fusion information entropy method of rolling bearing fault diagnosis based on n-dimensional characteristic parameter distance［J］．Mechanical Systems and Signal Processing，2017，88：123-136.
25	孙吉贵，刘杰，赵连宇. 聚类算法研究［J］．软件学报，2008，19（1）：48-61.
	SUN Jigui，LIU Jie，ZHAO Lianyu，Clustering algorithms research［J］．Journal of Software，2008，19（1）：48-61.
26	贾凡，严妍，张家琪. 基于K-means聚类特征消减的网络异常检测［J］．清华大学学报（自然科学版），2018，58（2）：137-142.
	JIA Fan， YAN Yan， ZHANG Jiaqi. K-means based feature reduction for network anomaly detection［J］．Journal of Tsinghua University（Science and Technology），2018，58（2）：137-142.
27	QIN H Y， ZHOU H P， CAO J W ．Imbalanced learning algorithm based intelligent abnormal electricity consumption detection［J］．Neurocomputing，2020，402：112-123.
28	GHADIRI S M E， MAZLUMI K. Adaptive protection scheme for microgrids based on SOM clustering technique［J］．Applied Soft Computing，2020，88：106062.
29	YAZDANI S， MONTAZERI G M. A novel gas turbine fault detection and identification strategy based on hybrid dimensionality reduction and uncertain rule-based fuzzy logic［J］．Computers in Industry，2020，115：103131/1-14.
30	杨占华，杨燕. SOM神经网络算法的研究与进展［J］．计算机工程，2006， 32（16）：201-202， 208.
	YANG Zhanhua， YANG Yan. Research and development of self-organizing maps algorithm［J］．Computer Engineering，2006，32（16）：201-202，208.
31	中华人民共和国住房和城乡建设部. 国家机关办公建筑和大型公共建筑能耗监测系统分项能耗数据采集技术导则［G/OL］．（2008-06-24）．.
32	胡振中，袁爽. 建筑能耗与环境监测系统标准化数据提取技术［J］．清华大学学报（自然科学版），2020，60（4）：357-364.
	HU Zhenzhong， YUAN Shuang ．Standardized data extraction techniques for building utility consumption and environmental monitoring systems［J］．Journal of Tsinghua University （Science and Technology），2020，60（4）：357-364．
33	YUN G Y， CHOI J， KIM J T ． Energy performance of direct expansion air handling unit in office buildings［J］．Energy and Buildings，2014，77：425-431．

文献	方法类别	数据降维特征选择方法	模式划分与基准建立方法	异常状态检测方法
［13］	监督学习	模糊C均值聚类	支持向量机	支持向量数据描述（SVDD）
［14］		深度置信网络	受限玻尔兹曼机	受限玻尔兹曼机
［15］		Pearson系数	三西格玛准则	CatBoost
［16］	无监督学习	主成分分析（PCA）	能源效率指标	关联规则
［17］		滑动窗口	自编码器	集成学习
［18］		K-均值聚类	分位阈值	分位阈值
［19］	混合监督学习	能耗平均值、日峰谷差率	DBSCAN/K-均值聚类	决策树算法
［20］		小波分析	组合神经网络	减法聚类算法
［21］		互补集合经验模态分解（CEEMD）	长短时记忆网络（LSTM）	Grubbs检验法
［22］		描述性时间序列特征	符号聚合近似	单类支持向量机

参数类别	参数名称	参数数量
设备参数	设备数量	24
设备参数	设备类别	4
运行参数	制冷主机运行参数	15
	冷冻水泵运行参数	2
	冷却水泵运行参数冷却塔风机运行参数	2 2
环境参数	室外相对湿度	1
环境参数	室外干球温度	1
其他参数	冷冻水供回水压差	1

主机数量	运行工况数量
主机数量	Ⅰ类	Ⅱ类	Ⅲ类	Ⅳ类	Ⅴ类
合计	14 611	14 559	36 544	37 181	2 575
单台	10 473	14 422	16 601	24 336	2 566
双台	4 138	137	19 943	12 845	9

系统工况	单主机				双主机
系统工况	制冷主机类	冷冻水泵类	冷却水泵类	冷却塔类	制冷主机类		冷冻水泵类	冷却水泵类	冷却塔类
Ⅰ类	［0.81，1.49］	［0，1.10］	［0，1.09］	［0，1.05］	［0.86，1.18］	［0.86，1.43］	［0.67，1.10］	［0.61，1.10］	［0.39，1.05］
Ⅱ类	［0.77，1.50］	［0，0.69］	［0，1.09］	［0，1.05］	［0.93，1.11］	［0.88，1.51］	［0.67，1.08］	［0.61，1.08］	［0.10，1.05］
Ⅲ类	［0.79，1.50］	［0，1.10］	［0，1.09］	［0，1.05］	［0.78，1.23］	［0.85，1.51］	［0.62，1.09］	［0.61，1.10］	［0.61，1.05］
Ⅳ类	［0.61，1.50］	［0，1.34］	［0，1.37］	［0，1.05］	［0.77，1.32］	［0.83，1.51］	［0.00，1.34］	［0.00，1.10］	［0.61，1.05］
Ⅴ类	［0.95，1.45］	［0，0.68］	［0，0.69］	［0，1.05］	［0.93，1.06］	［1.02，1.26］	［0.68，0.69］	［0.68，0.69］	［1.05，1.05］

数量	系统工况	少数类占总数比例/%	类内异常检测率/%
单台	Ⅰ类	1.87	97.96
	Ⅱ类	0.69	96.97
	Ⅲ类	3.09	96.49
	Ⅳ类	3.43	95.81
	Ⅴ类	0.19	100.00
双台	Ⅰ类	0.63	96.15
	Ⅱ类	0.73	100.00
	Ⅲ类	0.59	94.02
	Ⅳ类	0.82	93.33
	Ⅴ类	11.11	100.00