Accurate design cooling load is critical for rational designs of air-conditioning (AC) systems to ensure both indoor thermal comfort and energy efficiency. Coincident design day (CDD) enhances the accuracy of the design cooling load. In engineering applications, a small number of CDDs, applicable to numerous room parameter combinations, are preferred. Therefore, this study provides a reduced set of general CDDs to further enhance the practicality and convenience of CDD. A classification and optimization method is proposed to develop general CDDs for different room categories. The importance of room parameters is quantified to extract primary parameters for CDD classification. A multi-objective optimization is conducted to classify CDDs by maximizing CDD sharing ratios across room categories. The general CDDs are identified and comprehensively evaluated in terms of design weather parameters, design cooling load accuracy, energy-saving performance, and indoor thermal comfort. The results show that when general CDDs are used based on the window-to-wall ratio and solar heat gain coefficient, only one CDD is sufficient for the design cooling load calculation of a building. General CDDs can provide rational design weather data, thereby ensuring most design cooling load deviations within 8 % and reducing energy consumption by over 10 % Besides, using general CDDs achieves more acceptable thermal comfort, while using traditional design days can lead to overcooling. The general CDDs have exhibited potential for effective and expanded application in AC system designs.

High-altitude cold regions, such as the Tibetan plateau, face severe space heating demand, yet their low atmospheric density degrades air-source heat pump (ASHP) performance, and highly variable weather renders fixed control strategies ineffective. This study proposes a novel, data-driven adaptive control framework for a hybrid photovoltaic/thermal collector (PV/TC) and ASHP system, designed for a typical rural dwelling in Nagqu, Tibet. First, 254 heating-season days were classified into 17 distinct weather patterns using K-means clustering. Subsequent system simulation diagnosed three key operational dysfunctions inherent to conventional rule-based control: (i) frequent PV/TC on-off cycling due to irradiance fluctuations, (ii) excessive daytime ASHP operation on cold sunny days, and (iii) insufficient morning heating under extreme cold conditions. To address these, a weather-adaptive rule-based control strategy was developed, with its activation setpoints optimized for each pattern using a hybrid Particle Swarm Optimization-Hooke-Jeeves (PSO-HJ) algorithm. Results demonstrate significant performance improvements: PV/TC cycling was suppressed by 75-83% with minimal solar loss; daily heat collection increased by up to 28.5%, simultaneously reducing ASHP electricity consumption by up to 72.2%; and morning heating reliability was ensured under extreme cold, with a concurrent 6.9% increase in solar yield. This strategy's practicality lies in its reliance solely on standard weather forecasts and a pre-calculated lookup table of optimized setpoints, offering a transferable and computationally efficient solution for clean heating in high-altitude regions.

Urban building stocks are a dominant source of citywide carbon emissions and a primary target for urban decarbonization. Yet existing city-scale high-resolution studies still rarely provide planning-indicator-based and physically interpretable diagnoses of building-stock emissions, particularly with respect to realized demand and spatial heterogeneity. This study develops a bottom-up 1 km grid framework for Chengdu (534,000 buildings across 2102 grids) and establishes an interpretable scaling-law elasticity (SLE) model as the main analytical framework to relate grid-level total carbon emissions (TCE) to built scale, morphology, dominant function, building vintage, and an occupancy-informed utilization factor. The SLE baseline is then benchmarked against a geographically weighted random forest (GW-RF) with SHAP attribution to capture nonlinear responses and spatial non-stationarity. Results show that Chengdu's building-stock TCE reached 39.42 MtCO2 and exhibited a clear low-high-ultra-high regime structure. Hotspot formation was governed primarily by scale accumulation rather than height alone, while function, utilization, and vintage acted as systematic modulators. The ultra-highemission regime accounted for 10.8% of grids and delineated the priority zones for absolute emission reduction. The SLE model achieved strong interpretability with test R2 = 0.91, while GW-RF slightly improved predictive accuracy to 0.92 and further revealed nonlinear and spatially non-stationary driver effects through SHAP attribution. These findings provide an interpretable and planning-oriented basis for hotspot prioritization and differentiated carbon-control interventions at the intra-urban scale.

The thermal performance of building envelopes is important for effective passive solar heating, particularly in solar enrichment zones such as the Qinghai-Tibet Plateau. Large fluctuations in temperature and solar radiation pose significant challenges for passive solar heating. Therefore, building envelopes must simultaneously satisfy conflicting daytime and nighttime thermal requirements, which is difficult for conventional envelopes with static thermal performance. To address this challenge, this study proposes an adjustable window-Trombe wall composite envelope (AWT). The envelope integrates a vertically slidable window to simultaneously achieve daytime solar heat gain and nighttime insulation, and a Trombe wall with rotatable multilayer PCM-insulation louvers to enhance daytime solar absorption and nighttime heat release. The experiments were conducted in artificial controlled environment with different operating conditions to investigate the dynamic thermal performance of AWT. Simplified step-change boundary conditions for outdoor air temperature and vertical solar radiation were adopted to highlight the comparison of the effects of utilizing solar radiation. Compared with the non-adjusted condition, heat loss of the transparent part was reduced by 53.2% at night, while the heat loss of the opaque part was reduced by 85.6% and the indoor heat gain was improved by 1268.6 kJ·m-2 during nighttime. Overall, the solar heating efficiency increased from 22.7% to 72.6%. This study provides evidence for the application of adjustable building envelopes for passive solar heating in solar enrichment zone.

Traditional solar thermal systems with water as the working fluid are prone to both freezing and overheating, which impose operational challenges in cold regions. A solar-assisted heat pump integrating battery storage and radiant terminal energy system is proposed to meet the demands in well-insulated residential buildings. The heat pump converts the power generated by photovoltaic arrays into heat and further store in batteries and the building envelope. A test-bed in Shaanxi, China is employed to conduct full-time, 12-h, solar-based, and shutdown experiments, enabling a performance analysis of the operational characteristics and the building's thermal response. This study also develops a theoretical model for the proposed system and validates it using the measured data. Field experimental results show that the proposed system operates stably under cold-climate conditions. PV generation reaches 18.10 kWh/day and the high solar contribution is up to 83.0%. The system maintains indoor air temperature of seven rooms within 16.1-26.4 degrees C, even with 60 h of downtime, while the PMV values mostly fall within the Class I and Class II thermal comfort zones. Furthermore, the proposed system achieves the lowest carbon emissions among the compared energy systems. This verifies that the proposed system has significant research and application potential.

Variable Refrigerant Flow (VRF) systems in severe cold climates suffer from significant performance distortion due to high-frequency transient defrosting cycles. Mainstream building energy modeling (BEM) tools rely on rigid polynomial curves and static multipliers, failing to capture path-dependent heating debt and generating physically invalid extrapolations during extreme cold plateaus. To address this limitation, this study proposes a Physics-Informed Long Short-Term Memory (PI-LSTM) framework integrated with EnergyPlus via a timestep-level API co-simulation mechanism. By explicitly embedding Carnot theoretical limits and dew-point phase-change logic into the loss function, the PI-LSTM enhances thermodynamic consistency and mitigates the physics-blindness of purely data-driven models. Evaluated against field measurements in Harbin (−28°C), the PI-LSTM achieves high predictive fidelity (R2 = 0.945, CV(RMSE) = 6.58%), effectively suppressing physical boundary violations during capacity fluctuations. Under a single stochastically morphed worst-case design day realization, the co-simulation successfully propagates equipment-scale transient capacity deficits into zone-scale thermal responses. The simulation projects a severe maximum indoor temperature drop of 6.5°C, cross-verified by empirical indoor data with a dynamic sequence-level CV(RMSE) of 4.85%. To maintain continuous indoor thermal compliance—quantified as zero cumulative unmet degree-hours—this study identifies a case-specific equipment sizing multiplier of 1.4x for the evaluated hotel building configuration. Rather than providing a universal design rule, this research offers a transferable, physics-constrained simulation methodology for systematically quantifying transient thermal shocks in deep-freezing climates.

Located in the southern part of the North China Plain, Zhengzhou has experienced intensified winter temperature fluctuations in recent years due to global climate change. The alternation of cold waves and warm winters has increased the volatility of electricity demand, making trend prediction more challenging. To address this issue, this study develops a winter electricity demand forecasting model for Zhengzhou that incorporates extreme weather scenarios. The model is used to forecast electricity demand in Zhengzhou for January and February 2024. The results show that under normal, warmer-than-usual, and extremely cold scenarios, electricity demand in Zhengzhou increases by 12.5%, 6.6%, and 18.2%, respectively.

Rapid urbanization has intensified the urban heat island effect. Existing research has confirmed the association between urban form and land surface temperature, but the interaction and marginal effects of morphological factors on seasonal and diurnal variations of land surface temperature under different topographical conditions remain understudied. This study examines the seasonal and diurnal impacts of urban form on land surface temperature in cities with varying topographies, using Chengdu, Chongqing, and Guiyang as case studies. By integrating multi-source data to construct a multidimensional system and employing machine learning and Shapley additive explanation methods to analyze nonlinear relationships and interaction effects, the findings reveal，(1) Plains cities are moderated by built environment and landscape patterns, while mountainous cities are significantly influenced by topography and landscape patterns; plateau cities are more affected by topography and built environment; (2) Chengdu exhibits the strongest daytime heat island effect in summer, while Chongqing's high-temperature zones expand and contract seasonally. Guiyang maintains a south-hot, north-cool pattern during spring and summer daytime; (3) Among indicator interactions, the combination of high density of population and medium-to-high building density in summer leads to increased warming. This study provides support for developing targeted thermal environment control strategies and enhancing urban climate resilience in cities with diverse topographies.

The spatial distribution and temporal fluctuation characteristics of solar irradiance are of crucial importance for power generation in photovoltaic power plants and energy assessment in building-integrated photovoltaics. The all-sky radiance distribution can reflect sky conditions and is a significant factor influencing solar irradiance. However, existing studies lack temporal continuity in sky type classification and mostly focus on point data, making it difficult to support irradiance calculations for different orientations and tilted surfaces. Based on the observational data of the MS_321LR sky scanner in Xi'an from 2021 to 2022, this paper proposes a daily pattern matrix representation method for sky radiance, taking into account both temporal fluctuation and spatial distribution information. It also puts forward a sky classification method based on the K_means approach, namely the daily pattern sky classification method. Taking the VGAE_LSTM model as an example, a model framework of "sky classification + short-term radiance prediction" is constructed. On the basis of classifying the sky into four types: clear sky, circumsolar sky, parhelic sky, and meridional sky, the observational data set is reorganized and divided into four subgroups of data sets to train the prediction models respectively. Experiments show that the prediction accuracy of the prediction model with sky classification is improved by about 1 %-2 % compared with that of the unclassified model. Among them, the VGAE_LSTM model has the highest improvement amplitude, reaching 1.9 %. This model provides a new reference for solar irradiance, photovoltaic power generation, and meteorological research.

Evaluating thermal performance is pivotal for achieving Zero-Heating-Energy Buildings (ZHEB). Rather than targeting annual net-zero certification, this study adopts a worst-month screening criterion—zero auxiliary heating during the design heating month (January)—under monthly steady-state conditions. However, the established Solar Heating Fraction (SHF)-Solar Load Ratio (SLR) paradigm is hindered by computational complexity and lacks quantitative thresholds for ZHEB determination. To address these limitations, a hybrid framework combining Ordinary Least Squares (OLS) regression and a Four-Parameter Logistic (4PL) dose-response model is proposed. OLS-derived analytical equations link key design parameters to SHF, yielding a simplified coefficient-based screening tool for direct architectural application. The 4PL model characterizes the nonlinear SHF-SLR relationship and extracts region-specific ZHEB feasibility threshold descriptors within the explored design space, including the observed maximum performance response (Max SHF), performance evolution slope (Mid-slope), and the critical zero-auxiliary-heating threshold (SLR*). Building on the physical structure of SLR, a composite climatic index—the Radiation-Temperature difference Ratio (RTR)—is further proposed to guide design-strategy preferences across climates. Results reveal a climate-driven sensitivity gradient: in lower-RTR regions (e.g., Nagqu), SHF improvement depends more on thermal-loss minimization, particularly through reducing the north-facing window-to-wall ratio and envelope heat loss, whereas in higher-RTR regions (e.g., Lhasa), priorities shift toward solar-gain utilization, with length-to-depth ratio optimization and south-facing glazing becoming more prominent. Additional hourly benchmarking with EnergyPlus showed that the OLS predictions preserved the main city-wise performance ordering across climates. This study provides an operational evaluation framework and quantitative decision-support tools for ZHEB design in high-altitude, solar-enriched regions.

The design cooling load is a decisive factor in air-conditioning system design, with significant impacts on equipment sizing, energy use, and carbon emission. However, directly summing room-level design cooling loads (r-DCLs) generally produces an overestimated building-level design cooling load (b-DCL) due to the temporal misalignment. In this study, a refined theoretical method was proposed based on a process of temporal decomposition and reconstruction. Specifically, the r-DCL (calculated using the heat balance method) and temporal characteristics (extracted via Fourier-based frequency-domain analysis) are used to generate 24-hour hourly cooling load profiles for each room. These profiles are then temporally aggregated to obtain the building-level hourly cooling load profile. Finally, the maximum value of the aggregated profile is selected as the b-DCL. The results show that the MAE ranges from 3.10 W/m2 to 11.27 W/m2. The MAPE varies between 1.67% and 4.21%, while the MedAE falls within 2.60 W/m2 to 7.81 W/m2. Future work would focus on the development of an accurate r-DCL calculation method in the absence of sufficient meteorological data and the exploration of more robust statistical approaches for selecting DWD.

Protein thermal stability is fundamental to protein function, evolutionary adaptation and biotechnological applications. However, current predictors are limited by the noisy nature of optimal growth temperature (OGT) labels and the scarcity of experimentally determined melting temperatures (Tm), particularly for highly thermostable proteins. To address these challenges, we curated a unified dataset of similar to 90,000 proteins by integrating Tm values from high-throughput Thermal Proteome Profiling (TPP) measurements with OGT-derived high-stability sequences. This strategy increases thermal diversity while avoiding the introduction of OGT-associated noise into the low-stability class. Leveraging this dataset, we developed three feature-specific deep learning classifiers that separately combine global self-attention encoders with local convolutional residual modules, using sequence-evolutionary descriptors, ESM-2 contextual embeddings, or ProteinMPNN structural embeddings, respectively. Among these models, the ESM-2-based classifier - termed ThermoESM - achieved the strongest overall performance and exhibited excellent generalization on independent Limited Proteolysis-Mass Spectrometry (LiP-MS) datasets. Benchmarking against existing thermostability predictors further demonstrated ThermoESM's superior accuracy, along with balanced sensitivity and precision. Collectively, these results establish ThermoESM as a robust and widely applicable framework for large-scale thermal stability annotation and thermostability-guided protein engineering.

High penetration rate of renewable energy is critical for reducing building carbon emissions. In solar energy enrichment areas, it is technically feasible to harness solar energy via integrated building and energy systems to satisfy the heating and cooling demands of building. Outdoor design parameters are the basis for optimizing climate resource utilization and solar heating building design. However, conventional outdoor design parameters fail to adequately account for the hourly variability of outdoor climate conditions. It neglects the dynamic interaction between building performance and climate. Consequently, it is necessary to develop outdoor heating design days that can represent hourly variation characteristics relevant to solar energy utilization in buildings. Taking Lhasa City located on the Qinghai-Xizang Plateau as an example, this research establishes a dynamic heat transfer model to analyze indoor thermal environment responses under different building thermal parameters. The method for constructing outdoor heating design days is proposed using hierarchical clustering and evaluated by comparison with current outdoor design parameters using the Euclidean distance and heat consumption index. The results indicate that outdoor heating design days should include air temperature, horizontal global solar radiation, and global solar radiation on south facing. Compared with the conventional average day selection, outdoor heating design days derived from hierarchical clustering can better represent long-term weather data and reflect the adaptability of different building thermal performances to the climate. This approach improves heat consumption assessment from static estimation to hourly load profiling. It provides a robust basis for refined design and performance evaluation of solar heating buildings.

In recent years, China's renewable energy has developed rapidly, and photovoltaic buildings are an important form of renewable energy utilization. In order to conduct better preliminary energy-saving design and assessment for photovoltaic buildings, it is necessary to establish the meteorological year for photovoltaic buildings. Firstly, regarding the research on the radiation meteorological years of existing photovoltaic buildings, most of the studies focus on the global solar radiation as a meteorological parameter, while ignoring the two components of the global solar radiation when it reaches the horizontal surface: direct radiation and diffuse radiation. These two components have significant impacts on photovoltaic power generation under different weather conditions. Although there are a large number of existing studies on diffuse radiation estimation models, the current research does not take into account the influence of various factors in the atmosphere and the transparency of clouds. The accuracy of the diffuse radiation estimation model still needs to be improved. Second, the meteorological parameters involved in constructing the meteorological year for photovoltaic buildings all have strong temporal characteristics. In particular, direct radiation and diffuse radiation have the characteristics of rapid changes and high volatility. The existing methods for generating the radiation meteorological years are unable to meet the requirements of accurately capturing the temporal characteristics, rapid changes, and high volatility of these meteorological parameters. Therefore, this paper has accomplished the following works: (1)Based on the model of the diffusion fraction and diffusion coefficient after adding precipitation. By introducing cloud transmittance and atmospheric quality coefficient as input parameters, the accuracy of the existing diffuse radiation estimation model was further improved. (2)The method combining the autoregressive model and the long short-term memory network model (AR-LSTM) is adopted. The AR-LSTM model uses AR model to capture the short-term variation patterns in the radiation data, and LSTM model to capture the long-term patterns of solar radiation changes. Selecting the typical months from the long-term data that can represent the long-term average characteristics of meteorological factors such as direct radiation and diffuse radiation, thereby constructing the meteorological year for photovoltaic buildings. This paper conducts verification experiments using four cities as examples in the solar radiation zoning: Golmud and Lhasa in Zone Ⅰ, and Kashgar and Urumqi in Zone Ⅱ. The results show:①The R2 (coefficient of determination) values of the improved diffuse radiation estimation models for the four cities ranged 0.96–0.99, showing a good fit with the actual values; the values of RMSE (root mean square error) and MABE (mean absolute deviation error) were 0.007–0.021, indicating good prediction accuracy; the value of MBE (mean deviation error) was close to the actual values for all four cities. ②The fitting degree R2 of the various main meteorological parameters (dry bulb temperature, relative humidity, direct radiation and diffuse radiation) of the meteorological year for photovoltaic buildings in the four cities to the long-term average values ranged 0.78–0.99, indicating a good fit to the long-term average values. The AR-LSTM model shows better performance in terms of direct and diffuse radiation compared to Bi-LSTM, LSTM, RF, and XGBoost model. For other meteorological parameters, the relative error range of the AR-LSTM model from the long-term average is 0.02%-4.94%. The AR-LSTM model demonstrates a stronger advantage in maintaining the long-term statistical characteristics of the key radiation input variables for photovoltaic buildings.

To better design and optimize building envelopes and avoid undesirable heat losses and gains, simple yet effective methods for evaluating building energy efficiency are essential. This study develops a novel analytical method based on Green's function (GF) to calculate periodic heat transfer through building envelopes. Using GF, the analytical solution of one-dimensional temperature distribution in the frequency domain is derived by enforcing temperature and heat flux continuity at layer interfaces. The solution is inverse Fourier transformed into the time domain, yielding an explicit analytical expression formulated using simple functions. The interior surface temperature, heat flux, decrement factor, and time lag are determined by solving straightforward mathematical expressions. The method is also extendable to two-dimensional problems, demonstrating applicability to more complex scenarios. Validation and comparison results confirms that the GF method is both concise and accurate. Unlike conventional methods, such as the Laplace transform methods, the GF method is better suited for manual calculations and also avoids stability issues. In one-dimensional scenarios, the GF method results for interior surface temperature and heat flux agreed closely with the Kvalue software (based on the finite volume method), with maximum discrepancies of 0.056 degrees C and 0.293 W/m2, respectively. In twodimensional scenarios, it also matched finite difference simulations at the central region of each layer and the interior surface center of the composite structure, with a maximum temperature difference of 0.177 degrees C. The GF method provides a theoretical tool that combines physical clarity with engineering practicality for thermal performance and energy-efficient design of building envelopes.

The building sector accounts for 30%-40% of the global energy consumption, with office buildings being particularly significant because of their dense occupancy, abundant internal heat sources, and concentrated operational hours. Although extensive research has been conducted on building envelopes, a comprehensive review linking office building envelopes directly with thermal comfort is lacking. This study systematically analyzed the relationship between office building envelope components and indoor thermal comfort through a rigorous review of 186 peer-reviewed articles. The analysis revealed that the envelope components influenced the thermal comfort through three fundamental mechanisms of thermodynamic change: conduction, convection, and radiation. Advanced materials, such as aerogels, exhibit exceptional insulation performance, whereas phasechange materials integrated into walls reduce peak indoor temperatures. Smart technologies, including smart window systems and dynamic shading systems, can reduce the thermal discomfort hours. This study presents climate-specific optimization strategies: ultrahigh insulation for cold regions, efficient external shading for hot climates, and adaptable envelopes for moderate climates. This comprehensive review provides valuable insights for designers and engineers to optimize office building envelopes to enhance thermal comfort while maintaining energy efficiency.

Providing clean and cost-effective space heating for rural residences in Northern Shaanxi under cold climates remains challenging, and effective utilization of abundant solar resources is a promising solution. However, adaptability studies on solar heating systems tailored to this region are still limited. Meanwhile, conventional control methods often neglect both the differentiated heating demands across functional zones in rural residences and the thermal time-lag characteristics of indoor temperature. To address these gaps, this study proposes a method that combines system adaptability design with a differentiated control strategy based on model predictive control (MPC) for rural solar heating systems. Taking typical rural residential buildings in Northern Shaanxi, China, as a case study, an adaptability design of solar heating systems is conducted using economic and environmental benefits as evaluation indicators. To meet heterogeneous heating demands across functional areas, a Python-TRNSYS co-simulation-based MPC approach is developed, where TOPSIS combined with NSGAII balances economic performance, grid reliability, and thermal comfort. Compared with a constant-temperature control strategy and a time-of-use (TOU) price-based control strategy, the proposed method significantly improves living-room thermal performance. The percentage of time that indoor temperature remains within the comfort range increases by 20.20% and 9.78%, respectively. In terms of energy consumption, electricity use during peak periods decreases by 34.48% and 8.4%, while off-peak electricity use decreases by 2.55% and 3.12%, respectively. Meanwhile, PV self-consumption increases by 18.22% and 4.83%. Overall, the proposed method provides a feasible pathway for cost-effective and efficient winter heating in cold rural regions with abundant solar resources.

The Linpan settlements of the Chengdu Plain represent a distinctive rural landscape integrating nature, architecture, and vegetation, with significant implications for passive building performance and microclimate regulation. However, rapid urbanization has increasingly homogenized their spatial morphology, systematically eroding the vegetative buffering, shading, and natural ventilation that once reduced cooling energy demand by an estimated 2–3 °C. Traditional classification methods relying on qualitative field surveys are inadequate for capturing these energy-relevant spatial features at scale. To address this, we employ a deep learning approach using a convolutional neural network (CNN) built on the TensorFlow framework. A dataset of 2000 geo-tagged images (1600 training, 400 testing) was collected from 20 county-level regions. To manage multi-view imagery (aerial, street, and facade), we implemented a hybrid architecture with view-specific preprocessing and cross-view attention mechanisms. The baseline CNN achieved only 12% accuracy, reflecting high morphological similarity among Linpan types. By applying transfer learning with the Xception model, recognition accuracy improved to 31%. A confusion matrix further revealed probabilistic similarities between settlements, enabling objective, data-driven classification. Based on model output and remote sensing validation, we identify four spatial types: valley/riverside, plain, high-mountain, and hilly Linpan—each exhibiting distinct passive cooling potential, wind shelter efficiency, and solar access. The moderate accuracy (31%) not only reflects technical constraints but also signals ongoing spatial homogenization driven by urbanization, which diminishes the energy performance and thermal resilience of these vernacular landscapes. We conclude by proposing tailored protection and restoration strategies for each type, demonstrating that CNN-based analysis—despite current limitations—offers a replicable quantitative framework linking settlement morphology to energy performance, supporting climate-adaptive conservation and low-carbon rural development.