As the field of zero energy building design and research continues to progress, the use of data analysis methods is on the rise. These methods are applied to create assessment criteria, compare performance, and aid in design decision making. Decision trees, as a data-driven approach, offer interpretability and predictability, assisting designers in summarizing their design experience and serving as a foundation for design references. However, the current application of decision tree methods in the zero energy house sector primarily focuses on HVAC systems, lacking a comprehensive exploration from an architectural design perspective. Therefore, this study presents an empirical method for building and applying models based on decision trees, using zero energy house cases in severely cold regions of China as samples. Through an analysis of the interactions among various passive design parameters and the use of EnergyPlus for performance simulations, a decision tree model is established. This model aids in determining the recommended combinations of passive design parameters that meet the criteria of low energy consumption. Moreover, feature weighting highlights the most influential passive design parameters on building energy consumption, including the length of the architectural gestalt plane, the roof shape, and the ground thermal resistance. This research provides valuable methods and guidance for the design and construction of zero energy houses in severely cold regions of China.

The aim of this research was to investigate the thermal performance of Chinese vernacular skywell dwellings. This was done by the analysis of on‐site measurements of the environmental performance and computer simulation of solar conditions of eight vernacular skywell dwellings in three villages in southeastern China—Xidi, Zhifeng, and Yuyuan. Through the discussion of several heat inputs and outputs, it can be concluded that evaporative cooling had a substantial influence on the temperature in the skywells of dwellings in Xidi and Zhifeng villages. In addition to the buffering of temperature by thermal mass, evaporative cooling was found to further reduce the fluctuation in temperature inside the skywell and is likely to have been the main reason that the mean dry bulb temperatures inside the skywells in these villages were lower than the mean external dry bulb temperatures. Moreover, the construction materials of Chinese vernacular dwellings (skywell wall, floor material, wooden column, and panels) are likely to be hygroscopic.

To address the prevalent issues of high energy consumption, insufficient thermal comfort, and limited natural daylighting in rural houses, this study proposes a co-optimization method for thermal and daylighting performance based on a backpropagation neural network metamodel and the Non-dominated Sorting Genetic Algorithm III (NSGA-III) multi-objective optimization algorithm. A typical rural house in Suizhou was selected for field investigations and energy testing, with 1600 simulation scenarios conducted. Three key performance indicators—Energy Use Intensity (EUI), Predicted Mean Vote (PMV), and Useful Daylight Illuminance (UDI)—were extracted for rapid performance prediction using the metamodel. Multi-objective optimization was then performed with NSGA-III, and the Technique for Order Preference by Similarity to Ideal Solution method based on entropy weighting was applied to rank the solutions and identify the most balanced retrofit scheme. The results show that the optimized scheme reduced EUI by 22.47%, improved PMV by 44.66%, and increased UDI by 33.51%. The study identifies key factors influencing performance outcomes, including envelope thermal properties, window types, shading parameters, and sunspace configurations. Furthermore, compared with Multi-Objective Genetic Algorithm optimization conducted via Wallacei, the proposed approach demonstrates higher computational efficiency and more balanced Pareto-optimal solutions. The integrated “metamodel + multi-objective optimization” framework exhibits high predictive accuracy and practical applicability, providing an effective pathway to enhance both energy efficiency and indoor comfort in rural buildings.

As environmental concerns continue to rise, scholars worldwide are increasingly focusing on energy efficiency in rural energy use. This study explores the energy efficiency potential of rural households in China, aiming to identify the key factors that influence this potential through correlation analysis. By utilizing multiple regression analysis, the study investigates the relationship between these factors and household characteristics. The research is based on rural areas in central Jiangsu Province, China, specifically targeting near and far residents to assess variations in energy-saving potential across different household types. The research findings indicate that: (1) The residential energy-saving potential of near suburban residents is primarily influenced by PC, ER, PR and education, with PR having the most significant impact. (2) The residential energy-saving potential of far suburban residents is mainly influenced by ER, PR, and EC, with EC having the most significant impact. (3) For all residents, PR, EC, SE, and GP significantly influence their residential energy-saving potential, with PR and EC having the most substantial impact. Based on these observations, targeted policy recommendations are proposed.

Under carbon peaking and carbon neutrality goals, underground parking garages (UPGs) present distinct life-cycle carbon and energy performance characteristics, yet systematic life-cycle optimization remains limited. This study proposes an integrated framework that combines Life Cycle Assessment (LCA) and Multi-Objective Optimization (MOO) to simultaneously minimize life-cycle carbon emissions (LCCE), life-cycle cost (LCC), and operational energy (OE). Based on EnergyPlus simulations and Back-Propagation Neural Networks (BPNN), a high-accuracy surrogate model was developed to enable efficient Pareto optimization using NSGA-II. The TOPSIS method and Sobol global sensitivity analysis were further employed to rank optimal solutions and identify influential design variables. The results show that lighting power density (LPD) and air changes per hour (ACH) are the most influential factors, whereas envelope insulation thickness has low sensitivity due to soil thermal inertia. Compared with the baseline design, the optimal solution reduces LCCE, LCC, and OE by 17.65%, 5.48%, and 27.70%, respectively. The proposed framework provides insight into carbon reduction pathways in UPGs and offers quantitative decision support for their low-carbon design and operation.