Designing air-supported membrane structures often presents the challenge of balancing multiple performance objectives. To address this issue, this paper presents an integrated, multi-objective optimization framework. This framework automates the process from modeling and analysis to optimization by deeply coupling the parametric platform Grasshopper with the finite element software ANSYS and embedding a multi-objective evolutionary algorithm. The research focuses on developing a systematic parametric modeling method based on NURBS theory to precisely control complex surfaces and cable-net layouts. Various engineering applications have validated the effectiveness and versatility of the integrated framework. The integrated framework solves morphological analysis and parameter optimization problems for new structures, and it also applies to reverse-identifying the zero-stress state for existing structures. The results demonstrate that the integrated framework is an efficient, reliable, automated tool for solving performance-driven design problems in air-supported membrane structures.

Superhydrophobic coatings have demonstrated excellent defrosting performance. However, its defrosting performance when used in air source heat pumps (ASHPs) is still unclear. To promote its application in ASHPs, two ASHPs with same configuration were manufactured, one superhydrophobic and the other hydrophilic. Through the microscopic observation and experimental analysis, its defrosting performance under the different frosting conditions were revealed. Then, the effects of compressor speed and defrosting termination conditions on defrosting performance were quantitatively analyzed, respectively. Finally, according the analysis results, a new defrosting control strategy suitable for superhydrophobic ASHPs was proposed and validated. Experimental results showed that the superhydrophobic ASHP exhibited excellent defrosting performance under different frosting conditions. Compared to those of hydrophilic ASHP, its defrosting durations at 2/1 °C and −3/-4 °C were 75 s and 165 s shorter, and the defrosting energy consumption was reduced by 62.8% and 64.5%. However, as the defrosting control strategy of hydrophilic ASHP was used, the superhydrophobic ASHP encountered the excessive defrosting issue. When adopting the proposed defrosting control strategy, i.e. the compressor speed was 46 rps and defrosting terminated when the surface temperature reached 2 °C and lasted 30 s, the goal of minimizing the defrosting duration and energy consumption was realized.

During the defrosting operation of air source heat pump (ASHP), it stops supplying heating, or even absorbs heat from indoors. This results in the significant fluctuation of indoor thermal environment. However, the variation rule of indoor thermal environment during frosting-defrosting is still unclear, which severely restricts the heating quality improvement of ASHPs. To solve this problem, two test rigs, an air-water heat pump (AWHP) heating system and an air-air heat pump (AAHP) heating system, were built in an artificial environmental chamber. Through creating the frosting environmental condition, the practical space heating experiments of ASHPs during frosting-defrosting process were conducted. Then, the indoor thermal environment variation rules of the two ASHPs and their differences were analyzed. Results show that the indoor temperature during the defrosting process of both the AWHP and AAHP decreases obviously, and the decline amplitude in rooms using AWHP is greater. The temperature decline in rooms using AWHP is 2.99–4.66 °C, while 1.58–1.79 °C in the room using AAHP. For AAHP, during the later stage of its defrosting operation, the supply water temperature drops below the indoor air temperature, causing the indoor fan coil units to absorb heat from the indoor air instead of providing heating.

To maintain the high efficiency of air source heat pump (ASHP) under frosting conditions, the periodic defrosting is required. The earlier and later defrosting can both cause a decrease in its average coefficient of performance (COP) throughout the frosting-defrosting process. Accurately predicting the frosting rate and using it for defrosting control is a fundamental method to solve this problem. However, there is still lack of a universal model for predicting the frosting rate. To solve this problem, a mathematical model for variable-frequency ASHP was developed firstly, which can simulate its operating parameters under different conditions. Combining with the enthalpy diagram, a calculation method for the initial frosting rate of ASHPs was proposed. Then, an initial frosting rate prediction model suit for all operating conditions was developed utilizing the calculation results. Results indicate that the developed frosting rate prediction model for ASHPs demonstrates well predictive performance. The deviations between the predicted values and actual values are mainly within ±15

Heating terminal configurations affect the operating performance and indoor thermal response of air source heat pumps (ASHPs) significantly. However, most studies have focused on steady heating conditions, with limited consideration of the frosting-defrosting process. Consequently, their effects during frosting-defrosting cycles remain unclear, hindering improvements in ASHP heating performance for different terminals. To address this issue, an ASHP test rig was established in an artificial environment chamber, in which fan coil units (FCUs) and radiant floors (RFs) were connected to the same ASHP unit and operated separately. Then, experiments were performed under preset environmental conditions to compare the frosting-defrosting characteristics, heating performance, and indoor thermal environment of the ASHP system under the FCU and RF heating modes. Results showed that the defrosting will cause obvious fluctuations in supply and return water temperatures under both the FCU and RF heating modes, and the former is more significant. Compared to those under the RF heating mode, the frosting suppression performance under the FCU heating mode was better, with frosting durations prolonged by 21.7% and 29.8%, respectively. However, better heating performance was obtained under the RF heating mode. The average heating capacity and COP under the RF heating mode were 7.9% and 21.3% higher than those under the FCU heating mode. In addition, the defrosting had less effect on indoor thermal environment under RF heating mode. The indoor temperature and relative humidity kept nearly constant under RF heating mode, while fluctuated 3.13–4.02 °C and 6.03%–7.51% under the FCU heating mode.

During the defrosting operation of air source heat pumps (ASHPs), heating supply is interrupted. This leads to significant fluctuations in the indoor thermal environment, resulting in thermal discomfort for occupants. However, the dynamic behavior of indoor thermal conditions during the frosting-defrosting cycles remains unclear, which poses a major challenge to improving the heating performance of ASHPs. To solve this problem, two test rigs, one air-water heat pump (AWHP) heating system and one air-air heat pump (AAHP) heating system, were built in an artificial environmental chamber. Space heating experiments were conducted under controlled frosting conditions to investigate the indoor thermal environment during the frosting-defrosting cycles. By doing this work, the indoor thermal environment variation patterns during ASHP frosting-defrosting cycles were revealed for the first time. Results showed that the indoor temperature decreases notably during defrosting for both AWHPs and AAHPs, with greater declines observed for AWHPs. Specifically, the temperature drop ranged from 2.99 to 4.66 degrees C for AWHPs and from 1.58-1.79 degrees C for AAHPs. Furthermore, it was found that for both systems, the lowest indoor temperature does not occur at the end of the defrosting operation, as previously assumed, but rather 4.25-6.00 min later for AWHPs and 3.75-4.25 min later for AAHPs.

Windows play a decisive role in building energy consumption and indoor comfort. Dual-band electrochromic (DB-EC) glazing enables independent modulation of visible (VIS) and near-infrared (NIR) transmittance, and is typically operated in Bright, Cool, and Dark states to dynamically optimize environmental performance, offering promising potential for sustainable building applications. This study employs a co-simulation approach coupling EnergyPlus and Python to evaluate the performance of three DB-EC glazing types in a typical office building. A model predictive control strategy is designed to minimize energy consumption while maintaining visual comfort. Co-simulations are conducted across five major climate zones in China and three orientations. The DB-EC glazing types are compared against conventional glazing (EC, Low-E, and double glazing), with analyses focusing on state proportions, energy use (cooling, heating, and lighting), and useful daylight illuminance (UDI). Results indicate that three DB-EC glazing types exhibit an energy-saving difference of approximately 1 %-6%. In the Bright state, higher VIS and NIR transmittance provide better daylighting and solar heat gains, while in the Cool state, maintaining high VIS and blocking NIR helps reduce cooling loads without significantly increasing lighting demand. DB-EC glazing types achieve 1.0 %-22.9 % energy savings, outperforming EC (-3.7 %-21.0 %) and Low-E (0.8 %-13.8 %). Meanwhile, DB-EC glazing provides better daylighting comfort, with UDI500-2000 ranging from 41 % to 88 % compared with 40 %-75 % for EC glazing. Overall, DB-EC glazing performs particularly well in high-radiation, cooling-dominated climates, whereas its improvements are relatively limited in heating-dominated regions. These findings provide guidance for its further development and application.

Effective utilization of building energy flexibility is essential for balancing the power grid and provides economic benefits for building owners. Various flexibility sources have significant load-shifting potential and can participate more directly and effectively in demand response (DR). Nevertheless, there is a lack of a generalized comprehensive flexibility quantification method for grid-interactive building. To address this gap, this study proposes a novel dynamic-priority-based quantification method for comprehensive flexibility potential. Firstly, the physics-based flexibility quantification sub-models are separately established for individual flexibility sources. Secondly, the dynamic priority of flexibility sources is determined in real-time by the Technique for Order Preference by Similarity to Ideal Solution incorporated Entropy weight method. Finally, a comprehensive flexibility quantification method is developed based on the dynamic priorities and time windows of various flexibility sources. Validation results indicate that the proposed method can quantify the comprehensive flexibility potential of grid-interactive building accurately, with an average MAPE of 12.2% across various conditions. Additionally, the load characteristics and proportions of flexibility sources are comparatively analyzed. It indicates that utilizing multiple flexibility sources can achieve an average load shifting rate of 74% during DR periods, with active thermal energy storage, passive thermal energy storage, and appliance contributing 50.8%, 27.5%, and 21.7%, respectively. These findings provide valuable guidance for demand-side management and formulating DR dispatch strategies.

Superhydrophobic coatings exhibit excellent frost suppression performance. However, it remains unclear whether superhydrophobic ASHPs can maintain such performance during operation, given the complex influences of structural configuration and environmental parameters. In this study, based on a visualization test platform and an enthalpy difference laboratory, the microscopic frosting characteristics and frost suppression performance of superhydrophobic ASHPs were revealed. Furthermore, the frost suppression mechanism and failure modes were clarified through theoretical analysis. The experimental results showed that, compared to those of hydrophilic ASHP, its heating duration increased by 16.00

Gymnasium heating ensures stable indoor thermal environments for exercise during long winters in severe cold regions. However, the effect of different heating terminal types on exercisers' thermal responses is unclear. A winter field study in Changchun, China, was conducted to measure the thermal responses of subjects exercising on a treadmill at 5, 7, and 9 km/h (Ex 1, 2, and 3) in four spaces: radiator heating (RS), floor heating (FS), radiator and floor heating (RFS), and no heating (NS). The findings show that exercisers could hardly remain thermal neutrality in heated spaces but could achieve this state in NS at operative temperatures (top) of 16.8 °C and 14.8 °C during Ex 1 and Ex 2, respectively. RS (16.6 – 22.9 ℃) had a lower 80% acceptable top range than FS (18.1 – 25.1 ℃) and RFS (20.2 – 24.6 ℃). The lower top limit for 80% acceptability was 14.3 °C in NS. The 80% acceptable top ranges were 16.6 – 25.9 ℃, 15.5 – 24.8 ℃, and 9.6 – 21.8 ℃ during Ex 1, 2, and 3, respectively. In RFS, the preferred top was 22.1 °C during Ex 1, whereas it was below the measured range during Ex 2 & 3. In NS, it exceeded the measured range during Ex 1 and 2 and was 22.1 °C during Ex 3. This study improves our understanding of exercisers’ thermal responses in winter, indicates heating terminal energy‐saving potential in gymnasiums, and guides passive design of unheated sport spaces in severe cold zones.

During the defrosting operation of air source heat pump (ASHP) and a period thereafter, it stops supplying heating to indoor space. This leads to indoor temperature drop significantly, and further causes thermal discomfort for subjects. To improve the thermal comfort during the frosting-defrosting process of ASHP, two test rigs, one air-water heat pump (AWHP) heating system and one air-air heat pump (AAHP) heating system, were built in an artificial climate chamber. Under the standard frosting condition of 2/1 degrees C, space heating experiments were conducted to investigate the variations of indoor environment and thermal sensation of subjects during the frosting-defrosting process. Results showed that the indoor temperature declined notably during the defrosting operation for AWHP and AAHP. Specifically, the temperature decreased 2.6-4.6 degrees C for AWHPs, and 1.4-1.9 degrees C for AAHPs. Owing to the decrease in indoor temperature and the increase in draught sensation, the thermal sensation of subjects in AWHP-heated rooms and AAHP both decreased significantly. 78.4 % subjects in AWHPheated rooms and 73.5 % subjects in AAHP-heated rooms reported a downgrade in thermal sensation after defrosting. Then, the acceptable indoor temperature at the beginning time of defrosting were calculated, and they were respectively 20.74-25.09 degrees C and 19.91-24.45 degrees C for the AWHP-heated rooms and AAHP.

This study presents an experimental and analytical investigation into the mechanical behavior of PVDF-coated fabrics membranes with varying mass densities (650-1480 g/m2) and their corresponding weld seams. Eight types of industrially representative membranes from different manufacturers were selected for uniaxial tensile tests to record their load-displacement curves and ultimate tensile strength. Based on the measured tensile properties of each fabric, three different seam widths (from 20 to 110 mm) were designed for subsequent weld seam tensile testing. The comprehensive test program yielded a total of 140 valid membrane datasets and 430 valid weld seam datasets. Through analysis and comparison, a quantitative model was established to describe the mechanical performance of PVDF membrane weld seams under different mass densities and seam widths. The results clearly categorize the weld seam failure modes based on membrane strength of weld seams: low-and medium-strength membranes generally fail via direct tensile fracture, high-strength membranes exhibited peeling followed by tensile rupture, whereas ultra-high-strength membranes are predominantly subject to peeling failure, achieving only about 64 % of the membrane strength. Based on extensive experimental data, a predictive empirical model was developed that integrates multiple influencing factors-including membrane strength, seam width, and welding direction-to accurately estimate weld seam strength. This model demonstrates high reliability, with coefficients of determination (R2) reaching 0.99 in the warp direction and 0.98 in the weft direction, providing a practical tool for strength prediction in engineering applications.

Buildings participating in demand response (DR) programs are crucial for balancing supply and demand, particularly by leveraging thermal mass for load shifting without compromising comfort. However, existing energy flexibility quantification models are often complex and lack generalizability and real-world validation. This study addresses these issues by developing physics-based models to quantify the flexibility potential of building thermal mass. These models focus on the heat balance between indoor and outdoor environments and are validated using both measured and simulated data. Requiring only basic building information-such as thermal parameters, setpoint temperature, and outdoor meteorological data-the models avoid the need for complex operational data. Validation results show that over 90 % of flexibility potential predictions are within +/- 15 % of actual values, with an average error of 10.9 %. Compared to existing methods, the proposed model significantly improves predictive accuracy by 33 %. Additionally, the generalization capabilities of the models were tested with simulation data from various office models and climate zones, achieving a MAPE between 3.71 % and 11.76 %. These findings enhance the practical application of demand-side management in commercial buildings, contributing to more efficient and sustainable energy systems.

This paper proposes a semi-analytical method to address the challenges of geometric nonlinearity and the coupling of cable and membrane in the morphological analysis of air-supported membrane structures. Firstly, the main structure is simplified into a cylindrical membrane model based on the stress distribution pattern. Correction factors are then introduced to establish a large-deflection control equation that characterizes the discrete effects of the cable net and its stiffness contributions. To solve this equation efficiently, the sample space is constructed using a Latin hypercube sampling approach. When combined with the golden-section algorithm and symbolic regression techniques, an explicit functional relationship is derived between the correction factors and the key design variables. A comparison of the analytical algorithm implemented in MATLAB with the results of physical model tests demonstrates that this method reduces the total computation time to less than 2 min while ensuring computational accuracy. This significantly improves computational efficiency over traditional nonlinear finite element methods and provides a powerful tool for the rapid design of similar structures.

Frosting is inevitable for most air source heat pumps (ASHPs), making timely defrosting critical. Both rapid frosting and prolonged defrosting adversely affect power consumption and indoor thermal environments. To mitigate frosting impacts, most countries require the ASHP to undergo frosting-defrosting test before market entry. However, existing frosting test method fails to account for the outdoor environmental distribution and building heating load, resulting in the test results of low-temperature variable-frequency ASHPs (LVASHPs) varying significantly with the actual frosting performance. To address this issue, a mathematical model for LVASHP was developed first, and the operating parameters under various conditions were obtained. Then, using the calculation method of initial frosting rate, the frosting characteristics of LVASHPs in severe cold climatic region of China were analyzed, identifying the outdoor temperature range for the highest frosting rate. Combining the outdoor environmental distribution in the six provincial typical cities, the frosting test condition was determined. Finally, through analysis of the relative compressor speeds under frosting test condition, its setting method during frosting test was established. Results indicate that, when the building heating load is considered, the initial frosting rate first increases and then decreases as the outdoor temperature increases, with the maximum value existing. Based on the outdoor temperature range for the highest frosting rate and the temperature-humidity distribution during heating season jointly, the frosting test condition of LVASHP in severe cold climatic region of China should be -4 °C/85%. The compressor speed should be set to 55% or 60% of its rated speed during frosting test.

Reverse cycle defrosting is the mainstream defrosting technology for air source heat pumps (ASHPs). However, it is associated with issues such as poor indoor thermal comfort. To address these problems, a new defrosting method, hot gas direct pass defrosting, was proposed in this work. First, its working principle was introduced, and a control logic for it was proposed. Then, this new defrosting technology was applied to the super-hydrophobic and hydrophilic ASHPs. The stability, space heating performance and indoor thermal environment of the two ASHPs were investigated. Results indicate that when the hot gas direct pass defrosting is adopted, both the hydrophilic and superhydrophobic ASHPs operated stably during the 7 h experiments. During the hot gas direct pass defrosting process, the hydrophilic and superhydrophobic ASHPs could provide 32.66%-38.32% and 39.70% of the average heating capacity during the frosting process, respectively. The frosting duration and heating capacity of the superhydrophobic ASHP remained constant at 40 min and 4.03 kW, respectively. The total heating duration of the superhydrophobic ASHP was 18.02% longer, whereas the total defrosting duration was 77.78% shorter than that of the hydrophilic ASHP. Furthermore, the indoor temperature decreased notably during the frosting-defrosting process for both the superhydrophobic and hydrophilic ASHPs, with smaller temperature drop observed for the superhydrophobic ASHP. Specifically, the temperature drop ranged from 23.1 degrees C to 22.5 degrees C for the superhydrophobic ASHP and from 22.8 degrees C to 20.6 degrees C for the hydrophilic ASHP.

Model Predictive Control (MPC) has emerged as one of the most promising and extensively studied methods for implementing optimal control strategies in HVAC systems to enhance energy efficiency. However, the effectiveness of MPC critically depends on the accuracy of the underlying building model. Although both grey-box and data-driven models are widely used, each exhibits inherent limitations. To address these challenges, this study develops a Physics-Informed Neural Network (PINN) that integrates continuous-time ordinary differential equations (ODEs) for application in MPC within building energy systems participating in demand response programs. Firstly, a PINN incorporating first-order continuous-time ODEs for thermal modeling is developed to predict zone temperature. Compared to traditional grey-box and purely data-driven models, the PINN demonstrates improved accuracy in both one-step and multi-step temperature predictions. Furthermore, a PINN-based MPC strategy is proposed to optimize the modulation signal of a heat pump within the Building Optimization Testing (BOPTEST) framework. Results show that, during typical period, the PINN-based MPC enhances thermal comfort by 14.6 % to 31.9 %, while during peak period, improvements range from 65.2 % to 99.9 %. Regarding energy costs, the PINN-based MPC achieves reductions of 20.1 % to 29.3 % during typical period and 18.3 % to 19.9 % during peak period. Finally, this study examines the effects of the weight on physical knowledge and the amount of training data on thermal modeling performance. The findings demonstrate that integrating physical knowledge into neural networks significantly enhances prediction accuracy and control performance in building energy systems, especially with limited data, thereby improving the effectiveness of MPC.

Reducing the performance degradation of air source heat pumps (ASHPs) due to frosting during the heating season is critical. Previous studies have reported that a higher Characteristic Index for Configuration and Operation (CICO) improves the frosting suppression performance of ASHPs. However, selecting appropriate CICO for different regions is a challenge due to their varying frosting suppression needs. This study proposes two frost suppression grades (FSG 1 and FSG 2) to define acceptable levels of frosting loss, and develops a novel CICO design method based on the CICO-based equivalent temperature drop approach. Taking 161 high-humidity cities in China as examples, the study assesses the impact of different CICO on frosting losses, and determines the recommended CICO value for each region. Results indicate that increasing the CICO from 6 to 10 reduces frosting losses by an average of 37.5% in high-humidity regions, and raises the proportion of cities meeting frost suppression grades from 17% to 80%. To meet FSG 2, CICO values of 8 to 17 are generally sufficient, while achieving FSG 1 requires CICO values above 10 in 90% of cities and above 30 in 17 cities. These findings offer practical guidelines for the efficient application of ASHPs in high-humidity regions.