Restricted accessResearch articleFirst published online 2026
Comprehensive indicators for regional energy consumption of residential air conditioners—combining characteristics from thermal environment and cooling behaviour
Air conditioning energy consumption accounts for over 58% of total household energy usage. As the most widespread cooling devices, residential air conditioners (RACs) therefore play a critical role in urban energy systems. This study collected the operational data from 496 RACs via an IoT platform. Through correlation analysis, the SHapley Additive exPlanations (SHAP) and physical principles, the average duration time and outdoor air temperature were identified as the key indicators influencing the regional energy consumption of RACs. Subsequently, the devices were categorized into three distinct clusters by applying k-means clustering to their daily duration time and the night proportion. These clusters represent daytime (15%), night (61%) and whole day running pattern (24%). To integrate the effects of the thermal environment and occupants’ behaviour, two regional energy factors (REFIoT and REFClimate) were formulated using the key indicators of the duration time and temperature difference. These indicators were then used to develop regression models for predicting the regional energy consumption of RACs. The prediction models demonstrated strong performance, with those based on REFIoT achieving R2 values above 0.89. Through applying the framework and indicators to dataset of 2024, the performance was similar to the models of 2023.
United Nations Environment Programme. Frontiers 2025: The Weight of Time – Facing a new age of challenges for people and ecosystems. Nairobi: UNEP, 2025. https://wedocs.unep.org/handle/20.500.11822/48122.
3.
WeiWWuYZhaoJYanDAnJZhouX. Influence of AC supply modes with different levels of energy sufficiency on cooling energy consumption in urban residential buildings. Energy Build2025; 340: 115789.
4.
HuSYanDQianM. Using bottom-up model to analyze cooling energy consumption in China’s urban residential building. Energy Build2019; 202: 109352.
5.
DuJChenB. Indoor thermal environment and adaptive behaviors in summer of urban households in northern China: status and trend. Energy Build2024; 324: 114916.
6.
GuZHSunQWennerstenR. Impact of urban residences on energy consumption and carbon emissions: an investigation in Nanjing, China. Sustain Cities and Soc2013; 7: 52–61.
7.
Rama CurielJAThakurJ. A novel approach for direct load control of residential air conditioners for demand Side management in developing regions. Energy2022; 258: 124763.
8.
WangYWangZLiXLiHYangYXiongJ. Analysis on occupant behavior of split air conditioner in hotel buildings in the Yangze river region. Energy Build2022; 265: 112042.
9.
FrayssinetLMerlierLKuznikFHubertJ-LMilliezMRouxJ-J. Modeling the heating and cooling energy demand of urban buildings at city scale. Renewable Sustain Energy Rev2018; 81: 2318–2327.
10.
ShinMDoSL. Prediction of cooling energy use in buildings using an enthalpy-based cooling degree days method in a hot and humid climate. Energy Build2016; 110: 57–70.
11.
FengYDuanQChenXYakkaliSSWangJ. Space cooling energy usage prediction based on utility data for residential buildings using machine learning methods. Appl Energy2021; 291: 116814.
12.
LiXYaoR. Modelling heating and cooling energy demand for building stock using a hybrid approach. Energy Build2021; 235: 110740.
13.
DingYZhongCYangFKangZLiBDuanYZhaoZSongXXiongYGuoS. Low energy consumption thermochromic smart windows with flexibly regulated photothermal gain and radiation cooling. Appl Energy2023; 348: 121598.
14.
AnYChoiHKimEKimT. Experimental analysis of seasonal temperature characteristics and cooling and heating energy consumption of a slim double-skin window. Energy Build2022; 256: 111681.
15.
MengFZhangLRenGZhangR. Impacts of UHI on variations in cooling loads in buildings during heatwaves: a case study of Beijing and Tianjin, China. Energy2023; 273: 127189.
16.
DuanJLiNPengJLiuQPengTWangS. Clustering and prediction of space cooling and heating energy consumption in high-rise residential buildings with the influence of occupant behaviour: evidence from a survey in Changsha, China. J Build Eng2023; 76: 107418.
17.
Costa-CarrapiçoIRaslanRGonzálezJN. A systematic review of genetic algorithm-based multi-objective optimisation for building retrofitting strategies towards energy efficiency. Energy Build2020; 210: 109690.
18.
DengZChenYYangJChenZ. Archetype identification and urban building energy modeling for city-scale buildings based on GIS datasets. Build Simul2022; 15: 1547–1559.
19.
ZhaoXYinYZhangSXuG. Data-driven prediction of energy consumption of district cooling systems (DCS) based on the weather forecast data. Sustain Cities and Soc2023; 90: 104382.
20.
DingYZhangQYuanTYangF. Effect of input variables on cooling load prediction accuracy of an office building. Appl Therm Eng2018; 128: 225–234.
21.
MutaniGTodeschiVSantantonioS. Urban-Scale energy models: the relationship between cooling energy demand and urban form. J Phys: Conf Ser2022; 2177: 012016.
22.
SalataFFalascaSCiancioVCurciGGrignaffiniSde WildeP. Estimating building cooling energy demand through the cooling degree hours in a changing climate: a modeling study. Sustain Cities and Soc2022; 76: 103518.
23.
AydinalpMIsmet UgursalVFungAS. Modeling of the appliance, lighting, and space-cooling energy consumptions in the residential sector using neural networks. Appl Energy2002; 71: 87–110.
24.
PratavieraEZarrellaAMorejohnJNarayananV. Exploiting district cooling network and urban building energy modeling for large-scale integrated energy conservation analyses. Appl Energy2024; 356: 122368.
25.
XueKLiuMMaMHuMYanLChenXZengW. A framework for short-term energy consumption prediction based on room air conditioner group characteristics. J Build Eng2022; 52: 104400.
26.
LiXWenJBaiE-W. Developing a whole building cooling energy forecasting model for on-line operation optimization using proactive system identification. Appl Energy2016; 164: 69–88.
27.
YanLLiuM. A simplified prediction model for energy use of air conditioner in residential buildings based on monitoring data from the cloud platform. Sustain Cities and Soc2020; 60: 102194.
28.
ErfaniAJafarinejadTRoelsSSaelensD. Assessing the impact of the dataset’s size and quality on building thermal models for energy flexibility. Energy Build2024; 317: 114404.
29.
LiJ-YTangJ-ZZhaoX-ZFanBJiangW-YSongS-YLiJ-BChenK-DZhaoZ-G. A large-scale, high-quality dataset for lithology identification: construction and applications. Petroleum Science2025; 22: 3207–3228.
30.
TangZSunJTianYXuJZhaoWJiangGDengJGanX. CVRP: a rice image dataset with high-quality annotations for image segmentation and plant phenomics research. Plant Phenomics2025; 7: 100025.
31.
YangRHeGYinRWangGPengXZhangZLongTPengYWangJ. A large-scale ultra-high-resolution segmentation dataset augmentation framework for photovoltaic panels in photovoltaic power plants based on priori knowledge. Appl Energy2025; 390: 125879.
32.
MorkunaiteLPupeikisDTsalikidisNIvaskeviciusMManhangaFCCerneckieneJSpudysPKoukarasPIoannidisDPapadopoulosAFokaidesP. Efficiency in building energy use: pattern discovery and crisis identification in hot-water consumption data. Energy Build2025; 336: 115579.
33.
GB 50736-2012. Design Code for Heating, Ventilation and Air Conditioning in Civil Buildings. Beijing: The Standardization Administration of China, 2012.
34.
WuJLiuCLiHOuyangDChengJWangYYouS. Residential air-conditioner usage in China and efficiency standardization. Energy2017; 119: 1036–1046.
35.
SchoberPBoerCSchwarteLA. Correlation coefficients: appropriate use and interpretation. Anesth Analg2018; 126: 1763–1768.
36.
YangZChenYZhengZHuangQWuZ. Application of building geometry indexes to assess the correlation between buildings and air temperature. Build Environ2020; 167: 106477.
37.
LuYNiuDZhangSChangHLinZ. Ventilation indices for evaluation of airborne infection risk control performance of air distribution. Build Environ2022; 222: 109440.
38.
PereiraPFRamosNMM. Occupant behaviour motivations in the residential context – an investigation of variation patterns and seasonality effect. Build Environ2019; 148: 535–546.
39.
ChenCGaoZZhouXWangMYanJ. Energy consumption prediction and energy-saving suggestions of public buildings based on machine learning. Energy Build2024; 320: 114585.
40.
LinMChauCKHouHTangSK. Development of an interpretable machine-learning model for capturing nonlinear dynamics of multisensory interactions in public open spaces. Build Environ2025; 280: 113072.
41.
LiYGaoFYuJFeiT. Machine learning based thermal comfort prediction in office spaces: integrating SMOTE and SHAP methods. Energy Build2025; 329: 115267.
42.
LundbergSMLeeS-I. A unified approach to interpreting model predictions. In: Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc. 4–9 December 2017, Long Beach, USA, pp. 4768–4777.
43.
LiZMaJTanYGuoCLiX. Combining physical approaches with deep learning techniques for urban building energy modeling: a comprehensive review and future research prospects. Build Environ2023; 246: 110960.
44.
LiXWangXHeZChenXLiZ. Combining physical laws and ANN for predicting energy consumption of data center cooling systems. Energy Build2024; 311: 114170.
45.
YanLLiuMXueKZhangZ. A study on temperature-setting behavior for room air conditioners based on big data. J Build Eng2020; 30: 101197.
ChenXLiZDaiLZengWLiuM. Occupant heating patterns of low-temperature air-to-air heat pumps in rural areas during different heating periods. Builds2023; 13: 79.
48.
GawusuS. Evolving energy landscapes: a computational analysis of the determinants of energy poverty. Renewable Sustain Energy Rev2024; 202: 114705.
49.
CzétányLVámosVHorváthMSzalayZMota-BabiloniADeme-BélafiZCsoknyaiT. Development of electricity consumption profiles of residential buildings based on smart meter data clustering. Energy Build2021; 252: 111376.
50.
WangZCaoBLinBZhuY. Investigation of thermal comfort and behavioral adjustments of older people in residential environments in Beijing. Energy Build2020; 217: 110001.
51.
NodaLLimaAVPSouzaJFLederSQuirinoLM. Thermal and visual comfort of schoolchildren in air-conditioned classrooms in hot and humid climates. Build Environ2020; 182: 107156.
52.
ZhangYLuoXSunYLiuX. Impact of air-conditioning usage behavior of student groups on air-conditioning load simulation in university dormitories. Energy Build2024; 303: 113765.
53.
DaiLLiZChenXZhaiYFuCLiuZLiuM. Usage behavior characteristics of household air-conditioners during the extremely hot summer – A case study of Chongqing. Build Environ2023; 234: 110160.
54.
YanLLiuM. Predicting household air conditioners’ on/off state considering occupants’ preference diversity: a study in Chongqing, China. Energy Build2021; 253: 111516.
55.
GuXLiuMLiZLiuHChenXDaiL. Dynamic thermal demand analysis of residential buildings based on IoT air conditioner. Build Environ2023; 243: 110593.
56.
YanLLiJLiuMHuMXuZXueK. Heating behavior using household air-conditioners during the COVID-19 lockdown in Wuhan: an exploratory and comparative study. Build Environ2021; 195: 107731.
57.
SunYZhangCZhaoYLiJMaYZhuC. A systematic review on thermal environment and thermal comfort studies in Chinese residential buildings. Energy Build2023; 291: 113134.
58.
YangZXiaoHJiCWangBShiW. Effect of installation height of indoor unit on field heating performance of room air conditioner. J Build Eng2020; 32: 101527.
59.
XiY-A-MLiY-ZChenY-HJiangH-HHuangZ-B. Energy demand and carbon emission analyses of a solar-driven domestic regional environment mobile robot as household auxiliary heating and cooling method. Appl Energy2024; 371: 123743.
60.
ZhangMMaXWangWShengJCaoJChengZZhangX. Climate adaptation investments: short-term shocks and long-term effects of temperature variation on air conditioning adoption. Sustain Cities and Soc2024; 108: 105493.
61.
Bienvenido-HuertasDSánchez-GarcíaDPérez-FargalloARubio-BellidoC. Optimization of energy saving with adaptive setpoint temperatures by calculating the prevailing mean outdoor air temperature. Build Environ2020; 170: 106612.
62.
ZhanSChongA. Building occupancy and energy consumption: case studies across building types. Energy Built Environ2021; 2: 167–174.
63.
HanJHeZLiZMoNXieFYangWZhangC. Research on the coordinate regulation behavior of window opening and split air conditioners usage during summer in mixed-cooling houses in hot-humid regions. Energy Build2024; 322: 114710.
64.
XiaoHYangZShiWWangBLiBSongQLiJXuZ. Comparative analysis of the energy efficiency of air-conditioner and variable refrigerant flow systems in residential buildings in the Yangtze river region. J Build Eng2022; 55: 104644.
65.
DuYLiFKurteKMunkJZandiH. Demonstration of intelligent HVAC load management with deep reinforcement learning: real-world experience of machine learning in demand control. IEEE Power Energy Mag2022; 20: 42–53.
66.
WangJYik TangCSongL. Analysis of precooling optimization for residential buildings. Appl Energy2022; 323: 119574.
67.
RuanYMaJMengHQianFXuTYaoJ. Potential quantification and impact factors analysis of energy flexibility in residential buildings with preheating control strategies. J Build Eng2023; 78: 107657.
68.
WangJJiangYTangCYSongL. Analysis of predicted mean vote-based model predictive control for residential HVAC systems. Build Environ2023; 229: 109952.
69.
NorouziasasAAttiaSHamdyM. Impact of space utilization and work time flexibility on energy performance of office buildings. J Build Eng2024; 98: 111032.
70.
LiangJQiuYWangBShenXLiuS. Impacts of heatwaves on electricity reliability: evidence from power outage data in China. iScience2025; 28: 111855.
71.
LyuJLiJZhaoZMiaoXDuHLaiDYangYLianZ. How do people set air conditioning temperature setpoint in urban domestic–behavior model in Chinese three climate zones based on historical usage data. Energy Build2023; 284: 112856.
