BIM-Enabled Optimization of Thermodynamic Performance in Green Buildings

The primary objective of green buildings is to achieve energy-efficient operation and environmental compatibility throughout the entire life cycle. As a key parameter governing the interaction between buildings and their surrounding environment, thermodynamic performance directly influences heating, ventilation, and air conditioning (HVAC) loads and associated carbon emissions. BIM data encompasses comprehensive information on material properties of envelope structures, spatial configurations, orientation parameters, and other architectural attributes. By integrating the data into a multi-objective optimization model, systemic optimization of design-stage parameters—such as building form, window-to-wall ratio, and insulation material selection—can be conducted, enabling the generation of baseline design schemes that balance both energy efficiency and functional requirements.

A central challenge in the multi-objective optimization scheme is achieving systematic optimization of thermodynamic performance based on BIM data while simultaneously balancing energy efficiency, environmental benefits, and functional requirements. Thermodynamic performance in green buildings is influenced by multiple interrelated factors, including envelope structure, spatial configuration, and climatic adaptability. Although BIM technology enables the integration of geometric parameters, thermophysical material properties, and equipment operation data, significant difficulties remain in executing coordinated multi-objective optimization. The BIM data extraction workflow is illustrated in Figure 1. Overemphasis on energy consumption reduction alone may compromise the ability to regulate indoor thermal environments, while excessive prioritization of thermal comfort may diminish energy efficiency. Therefore, the development of a BIM-based multi-objective optimization framework is urgently required to resolve conflicts between competing performance objectives and to transition from fragmented parameter adjustments toward comprehensive system-level optimization.

1.png

Figure 1. Workflow of BIM data extraction

In the present study, the “multi-objective” scope includes four key dimensions. First, energy efficiency: this objective focuses on reducing the energy demands of HVAC systems by optimizing BIM-derived parameters such as the thermal transmittance of envelope components and building shape coefficient, thereby minimizing fossil fuel consumption. Second, carbon emission intensity: this objective constrains the optimization process within the bounds of life-cycle carbon emissions, promoting a balance between renewable energy utilization and envelope energy-saving measures to reduce greenhouse gas emissions such as carbon dioxide (CO₂) and methane (CH₄). Third, hygrothermal environmental performance: this dimension aims to regulate indoor air temperature, relative humidity, and surface temperature of envelope elements to mitigate condensation and mold growth risks, thereby enhancing building durability. Fourth, thermal comfort potential: this objective ensures that indoor thermal environmental conditions fall within the human comfort zone under scenarios involving natural ventilation and passive heating strategies based on the PMV index. These performance objectives are both interdependent and conflicting, necessitating the coupling of BIM data with advanced multi-objective optimization algorithms, generating a Pareto-optimal solution set through this process, thereby providing a scientifically grounded decision-making foundation for the optimization of thermodynamic performance in green buildings.

2.1 Multi-objective optimization overview

As illustrated in Figure 2, the multi-objective optimization of thermodynamic performance in green buildings based on BIM presents inherent conflicts among competing objectives. These conflicts are primarily concentrated in the dynamic trade-offs between energy efficiency, thermal comfort, and environmental benefits. For instance, a common contradiction arises between reducing building energy consumption and improving indoor thermal comfort. Increasing the thickness of insulation layers in envelope structures may reduce HVAC loads and subsequently lower energy use and carbon emissions.

2.png

Figure 2. Multi-objective optimization workflow of thermodynamic performance in green buildings based on BIM

However, such a strategy may simultaneously reduce the efficiency of natural ventilation, thereby impairing the occupants’ thermal comfort in terms of airflow perception. Similarly, enhancing natural daylight by increasing the window-to-wall ratio may improve visual and thermal comfort within indoor environments, yet it can also lead to excessive solar heat gain during the summer, raising cooling loads and intensifying carbon emissions. In addition, the pursuit of optimal hygrothermal conditions for indoor spaces may conflict with building durability objectives. Maintaining low indoor humidity to prevent mold may inadvertently increase the risk of surface condensation on envelope components, ultimately degrading material longevity. These contradictions stem from the intrinsic interdependence among the diverse performance requirements of green buildings. The coupling of envelope material properties and spatial configuration parameters within BIM data further exacerbates the complexity of coordinated multi-objective optimization.

To resolve the aforementioned conflicts, the proposed approach leverages the comprehensive data integration capabilities of BIM and the synergistic advantages of multi-objective optimization algorithms. By constructing a quantitative model encompassing objectives such as energy consumption, carbon emissions, thermal comfort, and hygrothermal transfer, geometric and material parameters extracted from BIM were transformed into computable objective functions. Intelligent algorithms were employed to generate a Pareto-optimal solution set, delineating the performance boundaries under various combinations of objectives. Furthermore, dynamic weighting strategies were applied to reflect the climatic context and functional requirements of specific green building scenarios, enabling personalized trade-offs across multiple objectives. Formally, let the optimization objective vector be denoted as b, and the total number of objectives be represented as V. The decision space is defined as A, consisting of M-dimensional decision vectors a. The feasible solution space is bounded by inequality constraints c_u(a)≤0 and equality constraints n_j(a)=0, with the lower and upper bounds of the search vector denoted as a_mlx and a_mlu, respectively. The multi-objective optimization problem considered in this study can be expressed as:

$\begin{aligned} & M I N b=d(a)=\left[d_1(a), d_2(a), \ldots, d_v(a)\right] \\ & v=1,2, \ldots V \\ & \text { s.t. } A=\left[a_1, a_2, a_3, \ldots, a_M\right] \\ & c_u(a) \leq 0, u=1,2, \ldots, l \\ & n_j(a)-0, j=1,2, \ldots, k \\ & a_{m i u} \leq a_m \leq a_{m l x}, m=1,2, \ldots, M\end{aligned}$          (1)

In multi-objective optimization, certain solutions located outside the Pareto front may exhibit reduced conflict among objectives compared to other types of solutions; such solutions can therefore be formally defined as follows:

(a) For a^*= [a^*₁,a^*₂,…,a^*_F] and a=[a₁,a₂…,a_F], if there exists a^*_m≤a_m and a^*_mp<a_mp, with m∈[1,M] and m_p=[1,M], then vector a^* is said to dominate vector a. Accordingly, for d(a^*) to dominate d(a), the following condition must be satisfied:

$\forall v, d_v\left(a^*\right) \leq d_v(a), v=1,2, \ldots, V$           (2)

(b) If a^* of the dominating vector a is a Pareto solution, within the Pareto solution set, no other feasible solution can dominate an optimal solution. Within the feasible domain, a solution a^* is defined as optimal if and only if there exists no solution a that satisfies condition d_v(a)≤d_v(a^*).

(c) If d(a^*) represents a globally optimal solution, then the condition d_v(a^*)≤d_v(a) must hold for all feasible solutions a.

2.2 Fundamental principle of the SSA

The optimization of thermodynamic performance in green buildings involves multiple interdependent objectives, including energy consumption, carbon emissions, thermal comfort, and hygrothermal transfer. These objective functions often exhibit nonlinearity and strong coupling. Furthermore, the high-dimensional solution space, shaped by BIM-derived variables such as envelope material properties and spatial configuration parameters, presents challenges for conventional algorithms, which tend to suffer from premature convergence or inadequate efficiency. SSA, inspired by the foraging strategies and anti-predation behavior of sparrow populations, combines global exploration capabilities with local exploitation efficiency. Through the role-based division of “discoverers” and “followers,” SSA effectively balances population diversity with convergence speed. This balance is especially advantageous in resolving conflicts among energy efficiency, thermal comfort, and low-carbon objectives, as it enables the dynamic adjustment of step sizes and population positions to identify Pareto-optimal solutions across different objective combinations. From a BIM-driven optimization perspective, SSA exhibits strong compatibility with engineering scenarios due to its adaptive parameter adjustment mechanism. BIM-based models integrate comprehensive architectural datasets, forming a complex multi-input–multi-output optimization system. The adaptive weight update strategy employed in SSA enables the handling of nonlinear relationships inherent in BIM data. For example, in cold climate regions, the algorithm can automatically increase the influence weight of envelope insulation parameters on energy-related objectives while concurrently satisfying constraints defined by the indoor PMV index for thermal comfort.

Moreover, SSA’s built-in “alert mechanism” reflects the threshold control requirements common in practical engineering applications. By introducing a safety distance operator, the algorithm ensures that all solutions remain within feasible domains defined by green building standards, thereby avoiding the risk of obtaining “theoretically optimal but practically infeasible” solutions—a common drawback in traditional optimization algorithms. Formally, let the current iteration number be denoted as s, and the maximum number of iterations be represented as IT_MAX. The position of the u-th sparrow in the k-th dimension is expressed as A_u,k. A random number is defined by β∈(0,1], while the early warning value is represented by E₂∈(0,1], and the safety threshold by TS∈(0.5,1]. The matrix m×f consists of elements all equal to 1. The stochastic term W represents a random number that follows a normal distribution. The position update rule for the sparrow population in SSA is given as:

$A_{u, k}^{u+1}=\left\{\begin{array}{l}A_{u, k}^s \cdot \exp \left(\frac{-u}{\beta \cdot I T_{M A X}}\right), E_2<T S \\ A_{u, k}^s+W \cdot M, E_2 \geq T S\end{array}\right.$          (3)

Let the optimal position of the current discoverer be denoted by A_o, and the globally worst position by A_WO. A matrix of dimension 1×f is represented by X, where X⁺=X^T(XX^T)^-1. The position update formula for joiners is expressed as:

$A_{u, k}^{s+1}=\left\{\begin{array}{l}W \cdot \exp \left(\frac{A_{W O}^s-A_{u, k}^s}{u^2}\right), u>\frac{V}{2} \\ A_o^{u+1}+\left|A_{u, k}^s-A_o^{u+1}\right| \cdot X^{+} \cdot L \text { other }\end{array}\right.$          (4)

Let the global best position be denoted by A_BE, the step-size adjustment coefficient by α, and a uniformly distributed random number within the interval by j∈[-1,1]. The fitness value of the current sparrow is denoted by d_u, the global best fitness value by d_h, and the global worst fitness value by d_q. A small positive constant γ was introduced to prevent division by zero. The position update formula for the vigilant group is given as:

$A_{u, k}^{s+1}=\left\{\begin{array}{l}A_{B E}^s+\alpha \cdot\left|A_{u, k}^s \cdot A_{B E}^s\right|, d_u>d_h \\ A_{u, k}^s+j \cdot\left(\frac{\left|A_{u, k}^s-A_{B E}^s\right|}{\left(d_u-d_q\right)+\gamma}\right), d_u=d_h\end{array}\right.$           (5)

Steps of SSA are as follows:

Step 1: Population initialization based on BIM data

During the population initialization phase, core variables relevant to thermodynamic performance optimization in green buildings were first extracted from the BIM-based model, forming a high-dimensional decision variable space. Variable ranges were determined in accordance with green building design standards and climatic region characteristics. For instance, the window-to-wall ratio is typically constrained within the range of 0.2–0.7, and the upper and lower bounds of insulation thickness are defined based on regional energy efficiency codes. Simultaneously, the population size and maximum number of iterations were initialized. Proportions for the discoverer group and joiner group were also set, with the remaining individuals designated as members of the vigilant group. Through the structured output of BIM data, physical building parameters were transformed into a machine-readable initial population, ensuring that the optimization variables precisely correspond to the actual design elements of green buildings.

Step 2: Multi-objective fitness evaluation and non-dominated sorting

Using climate data, thermal material properties, and simulation tools such as EnergyPlus integrated within the BIM model, the multi-objective fitness values of each individual in the initial population were computed, including building energy consumption, life-cycle carbon emissions, the PMV index for thermal comfort, and a condensation risk coefficient. A non-dominated sorting algorithm was applied to rank individuals, yielding a Pareto front solution set comprising “non-dominated solutions” that demonstrate superior performance across multiple objectives. This step translates the abstract performance goals of green buildings into quantifiable fitness functions through BIM data-driven simulation calculations, establishing an evaluation baseline for subsequent search.

Step 3: Position update of discoverers and global exploration

Discoverers, representing the individuals with superior fitness values in the population, are responsible for global exploration. In the context of green building optimization, discoverers prioritize adjustments to variables that exert significant influence on multiple objectives, such as window-to-wall ratio and building orientation. For instance, in regions characterized by hot summers and mild winters, discoverers may increase the external window shading coefficient to reduce summer heat gain, while moderately decreasing the window-to-wall ratio to maintain adequate daylighting. During the position update, green building code constraints were embedded to ensure engineering feasibility, thereby preventing the generation of “theoretically optimal but non-compliant” solutions.

Step 4: Position update of joiners and local exploitation

Joiners, who follow discoverers in the population hierarchy, engage in fine-grained local search through competitive mechanisms. Within the green building context, joiners concentrate on the neighborhoods of high-quality solutions identified by discoverers. For example, in response to a specific combination of high-performance insulation materials, the joiners refine the relationship between insulation thickness and the thermal conductivity of wall materials or optimize the angle of shading elements to balance winter heat gain and summer shading. This step enhances the quality of solutions within targeted objective combinations and increases the density of the Pareto front.

Step 5: Position update of vigilant individuals and threshold constraints

Vigilant individuals are tasked with monitoring whether the population is approaching “risk zones.” It is triggered when the early warning value exceeds a defined safety threshold. In green building optimization, risk zones may correspond to scenarios such as excessive condensation risk or unacceptable levels of carbon emission intensity. For instance, if an overly large window-to-wall ratio results in a sharp increase in cooling energy demand during summer, vigilant individuals enforce corrective actions by constraining the window-to-wall ratio within regulatory limits and synchronously adjusting shading coefficients or glazing types. This step ensures that the solutions remain within the feasible performance domain defined by green building standards and prevents safety boundary violations that may occur due to over-optimization toward a single objective.

Step 6: Multi-objective performance revalidation and position correction

Following the position updates of discoverers, joiners, and vigilant individuals, a revalidation of multi-objective performance was conducted using BIM data and associated simulation tools for all individuals in the population. Specifically, updated envelope parameters were exported from the BIM-based model and input into EnergyPlus to recalculate energy consumption and carbon emissions. Indoor thermal conditions and the PMV index were evaluated using Daysim, while WUFI was employed to assess hygrothermal transfer and condensation risk. If performance anomalies are detected due to parameter adjustments, corrections are applied to the corresponding individual’s position based on the revalidation results, ensuring that each solution retains physical significance and logical consistency with the BIM-based model data. Through this closed-loop process of “simulation–validation–correction,” the mapping accuracy between algorithm-generated solutions and the actual thermodynamic performance of green buildings can be significantly enhanced.

Step 7: Termination criteria and output of optimal solution set

The optimization process was terminated upon satisfying one of two conditions: reaching the maximum number of iterations or achieving convergence in the Pareto front solution set. Once a termination criterion is met, the resulting Pareto-optimal solution set is output, encompassing energy consumption, carbon emissions, thermal comfort, and hygrothermal performance objectives. Each solution in the output corresponds to a specific set of BIM parameter configurations. Project teams may apply the Analytic Hierarchy Process (AHP) to assign weights to the various objectives based on project-specific priorities, selecting optimal schemes from the solution set, thereby yielding the final green building thermodynamic performance optimization scheme that achieves a balance among energy efficiency, thermal comfort, and regulatory compliance.

Figure 3 illustrates the complete workflow of SSA.

3.png

Figure 3. Workflow of SSA

2.3 Parametric-driven building performance modeling

The core foundation of BIM-driven thermodynamic performance optimization for green buildings lies in the integration and parametric mapping of all architectural elements enabled by BIM technology. As a digital carrier, the BIM-based model precisely encodes geometric parameters, thermophysical material properties, climatic boundary conditions, and functional requirements, forming a multi-dimensional parametric database. Through parametric modeling, physical characteristics such as envelope component configuration and spatial layout are translated into quantifiable input variables. Concurrently, thermodynamic performance objectives—such as energy consumption, thermal comfort, and carbon emissions—are defined as output variables, establishing a functional input–output mapping relationship. For example, modifications to the window-to-wall ratio within the BIM-based model are programmed to automatically trigger energy simulation modules to recalculate heating and cooling loads, while simultaneously prompting the thermal comfort simulation module to assess changes in the PMV index. This mechanism enables a digital pre-evaluation of building performance, forming a data-driven logic framework that supports data-oriented optimization. Figure 4 illustrates the workflow of parametric-driven building performance modeling.

4.png

Figure 4. Workflow of parametric-driven building performance modeling

The parametric-driven mechanism based on BIM interprets the intrinsic relationships between parameter variables and thermodynamic performance objectives through computational simulation technologies, which is a critical principle in optimization. Initially, key sensitivity parameters that exert significant influence on the thermodynamic performance of green buildings—such as the thermal transmittance of envelope structures, building shape coefficients, and phase change material (PCM) transition temperatures—were identified. The impact of parameter variation on multiple objectives was quantitatively assessed through orthogonal experiments or response surface methodologies. For example, in buildings located in cold climate regions, an increase of 10 mm in insulation thickness may result in a 5%–8% reduction in energy consumption, while concurrently causing a 3%–5% decrease in natural ventilation efficiency—exemplifying a typical objective conflict scenario. Subsequently, a multiphysics coupling model incorporating principles of heat transfer, fluid dynamics, and meteorology was constructed. Through BIM data interfaces, real-time transfer of parameter variables across various simulation tools was achieved, enabling accurate simulation of thermal flux distribution, airflow patterns, and the evolution of indoor thermal environments resulting from parameter adjustments.

To address the inherent conflict among thermodynamic performance objectives in green buildings, an intelligent optimization algorithm—driven by BIM data—operates through a closed-loop mechanism comprising parameter perturbation, performance evaluation, and solution set iteration, thereby enabling coordinated optimization. Specifically, the parametric output variable space generated by the BIM-based model was first transformed into a solution space interpretable by intelligent algorithms. SSA was then employed to emulate biological group cooperation and Pareto dominance relationships, facilitating the identification of non-dominated solutions that simultaneously satisfy multiple objectives. For instance, SSA discoverer individuals are responsible for global exploration of optimal combinations between window-to-wall ratio and insulation thickness. Joiner individuals refine local solutions in high-quality regions by adjusting shading component angles and glazing specifications. Vigilant individuals enforce boundary constraints to prevent violations of condensation risk thresholds or code limitations. Ultimately, a Pareto-optimal solution set was generated, encompassing energy consumption, carbon emissions, and thermal comfort objectives. Further, with the support of BIM’s visualization and data traceability capabilities, engineering feasibility verification of the solution set was conducted. For example, envelope structure parameters corresponding to an optimized solution can be reverse-mapped into the BIM-based model to verify whether node configurations comply with construction codes and whether thermal properties align with design parameters. Through this integrated technical chain—comprising data modeling, intelligent optimization, and engineering validation—optimization outcomes of thermodynamic performance are ensured to be not only multi-objective compliant but also practically implementable throughout the full life cycle of green building projects.

Table 1 presents the parameters used in the constructed BIM-based model. The dataset was organized across four principal dimensions: geometric, thermal, operational, and climatic. In terms of geometric parameters, the number of floors (6), building orientation (south), and floor height (2.9 m) directly influence the spatial form and daylighting/ventilation potential of the building. These constitute the fundamental geometric framework for constructing the physical building model. For thermal parameters, values are assigned for the thermal transmittance of external walls (0.114), roof (0.097), and windows (3.562). The significant variation among these values highlights the differing insulation performance of the envelope components. In particular, the relatively high thermal transmittance of windows may render them critical interfaces for heat transfer. These parameters serve as key inputs for energy consumption analysis and hygrothermal performance evaluation and are directly linked to the impact of envelope optimization on thermodynamic performance. The operational parameters and climatic parameters further refine the usage scenarios and external environmental conditions. Operational parameters, including an average equipment power density of 0.15, indoor cooling and heating set-point temperatures of 25℃ and 18℃ respectively, and a total occupancy of 71, reflect the internal thermal loads and usage conditions. These influence the dynamic behavior of indoor heat gain and loss. Climatic parameters are characterized by 189 cooling degree days, 168 heating degree days, an annual mean dry-bulb temperature of 15.6℃, and an annual mean globe temperature of 11.2℃. These values clearly define the local climatic context and provide the boundary conditions necessary for dynamic thermal simulation and the PMV thermal comfort index calculation. The integration of these parameters enables the BIM-based model to simulate thermodynamic performance across varied climatic regions and operational conditions, while also serving as a robust basis for verifying the indoor thermal comfort compliance rate.

As shown in Table 2, the proposed method achieved a mean Inverted Generational Distance (IGD) value of 1.12E-03 and a standard deviation of 2.36E-02 in Sample 1, both of which were lower than those produced by Strength Pareto Evolutionary Algorithm 2 (SPEA2) (2.89E-03, 3.89E-03), Multi-Objective Cuckoo Search (MOCS) (1.21E-02, 3.56E-02), and Pareto Archived Evolution Strategy (PAES) (2.25E-03, 1.54E-03), indicating a superior central tendency in the proposed approach. In Sample 2, the proposed method recorded a mean of 8.45E-03 and a standard deviation of 7.79E-03. MOCS recorded a mean of 5.13E-02 and a standard deviation of 5.89E-1. This shows that the method proposed in this study has a better degree of data dispersion. In Sample 3, the proposed method again demonstrated superior convergence, with a mean value of 1.36E-03—lower than both MOCS (4.12E-02) and PAES (1.62E-03)—along with a reasonable standard deviation of 2.56E-02. Comparable advantages were also observed in Samples 4 and 5. For instance, in Sample 4, the proposed method yielded a mean IGD value of 1.14E-03, significantly lower than MOCS (5.17E-02), further reflecting the robustness of the method.

Table 1. BIM-based model parameters

Table 2. IGD test results for the multi-objective optimization method for green building thermodynamic performance

Table 3. Spatial evaluation results of the multi-objective optimization method for green building thermodynamic performance

Table 4. Selected parameters of non-dominated feasible solutions from the multi-objective optimization framework for green building thermodynamic performance

From the results presented in Table 2, it can be concluded that the proposed BIM-driven multi-objective optimization method for green building thermodynamic performance demonstrates excellent performance in IGD testing. The relatively low mean values suggest that the method can more accurately approximate the optimal solution, indicating high optimization efficiency. The relatively small standard deviation values indicate strong solution stability and minimal dispersion, further validating the reliability of the approach.

As shown in Table 3, the proposed method achieved a mean value of 3.48E-03 and a standard deviation of 4.56E-02 in Sample 1, outperforming SPEA2 (8.26E-03), MOCS (6.45E-03), and PAES (4.56E-03) in terms of convergence toward optimal solutions. In Sample 2, the proposed method yielded a mean of 3.45E-02 and a standard deviation of 7.36E-02, demonstrating a clear advantage over SPEA2, which reported a mean as high as 6.58E-02. In Sample 3, the mean and standard deviation achieved by the proposed method were 4.15E-03 and 1.17E-03, respectively, both considerably lower than those of MOCS (4.28E-02, 1.26E-02). In Samples 4 and 5, superior accuracy and robustness were also observed. For example, in Sample 4, the proposed method yielded a mean of 2.36E-03, which was notably lower than that of SPEA2 (4.78E-03), confirming both precision and consistency. The data in Table 3 validate the effectiveness of the proposed BIM-driven multi-objective optimization method for thermodynamic performance evaluation in spatial contexts. The relatively smaller mean values indicate an enhanced ability to concentrate solutions near the optimal region, highlighting the accuracy and efficiency of the BIM-integrated optimization framework. The deep coupling of intelligent algorithms with the BIM-based model facilitates the coordinated optimization of envelope parameters and spatial geometry. The stable standard deviation values reflect the method’s reliability, with low result dispersion across test instances.

As shown in Table 4, under the energy-oriented objective, the annual energy consumption is 2.3256×10⁷ kWh, with a PMV of 1.68, a south-facing window-to-wall ratio of 30%, and envelope configurations consisting of glass type 5 and wall type 1. Under the combined energy and comfort objective, the annual energy consumption increases to 2.78512×10⁷ kWh, with an improved PMV of 1.23. The window-to-wall ratio is increased to 60% (still south-facing), the orientation is adjusted to southeast, and envelope materials are updated to glass type 1 and wall type 2. Under the comfort-oriented objective, energy consumption further rises to 3.4526×10⁷ kWh, with the PMV reduced to 0.98. A larger window-to-wall ratio (65%) facing north and a southeast-south orientation are adopted, along with glass and wall materials both classified as type 3. The distinct variations in parameter settings across the three objectives highlight the inherent trade-offs among competing performance goals during multi-objective optimization. These results demonstrate the practical value of the BIM-driven optimization framework developed in this study. By integrating multiple objectives—including building energy demand and indoor thermal comfort potential—and coupling intelligent algorithms with the BIM-based model, coordinated optimization of envelope parameters and spatial morphology was achieved. For example, in the energy-prioritized scenario, the window-to-wall ratio was minimized to reduce energy loads, whereas in the comfort-prioritized scenario, adjustments to window ratio and building orientation were made to enhance the PMV index. These parameter configurations serve as specific input conditions for subsequent thermal comfort compliance validation. During the validation phase, thermal parameters exported from the BIM-based model can be used to reverse-adjust the optimization model, ensuring that performance targets are met under different objective priorities.

As illustrated in the temperature simulation graph in Figure 5, air temperature (blue line) exhibited pronounced seasonal fluctuations, beginning at lower levels in February, rising progressively to a peak in mid-year, and then gradually declining. This variable exhibited the widest range of variation. In contrast, mean radiant temperature (red line), effective temperature (green line), and outdoor dry-bulb temperature (purple line) remained relatively stable, consistently fluctuating around approximately 20℃. This stability indicates that the proposed optimization scheme contributed to effective regulation of the indoor thermal environment. In the corresponding humidity simulation, relative humidity (blue line) displayed dynamic monthly variation, ranging from a minimum of approximately 35% to a peak exceeding 80%. This variability reflects the seasonally influenced characteristics of indoor humidity and highlights the susceptibility of indoor moisture conditions to external climatic changes. These simulation results demonstrate the effectiveness of the BIM-driven thermodynamic optimization framework in improving indoor thermal conditions. The observed stability in radiant temperature and effective temperature confirms the positive impact of optimized envelope parameters on indoor thermal regulation, aligning with the framework’s goal of enhancing hygrothermal performance through multi-objective optimization. Furthermore, the seasonal dynamics of humidity provide critical input for evaluating the thermal comfort compliance rate. By incorporating both temperature and humidity indicators, a more comprehensive understanding of their combined influence on thermal comfort can be achieved. These results also offer a data-driven foundation for reverse adjustments to the thermodynamic optimization model. Consequently, the framework’s ability to enhance indoor thermal comfort potential under varying climatic conditions and usage scenarios can be further validated and refined.

5-1.png

(a) Temperature

5-2.png

(b) Humidity

Figure 5. Indoor thermal comfort simulation for the BIM-driven thermodynamic performance optimization scheme in green buildings

Table 5. Simulated unit-area energy consumption results for different glass types in green buildings

Table 6. Simulated unit-area energy consumption for different window-to-wall ratios in green buildings

As shown in Table 5, the simulated unit-area energy consumption for various glass types used in green buildings revealed distinct differences. In terms of total energy consumption, Low-E glass exhibited the lowest value (135.236), followed by insulated glass (138.235), vacuum glass (138.256), smart glass (139.586), and coated glass (138.625). The air conditioning energy consumption values were recorded as 62.365 for Low-E glass, 57.236 for insulated glass, 56.321 for vacuum glass, 57.235 for smart glass, and 56.215 for coated glass. It was observed that Low-E glass exhibited the lowest total energy consumption among all tested materials. In contrast, vacuum glass and coated glass demonstrated relatively lower air conditioning energy demands. These results indicate that significant differences in both total and air conditioning-related energy consumption are associated with variations in glass material properties. These findings highlight the critical role of BIM-driven optimization of envelope parameters in green building design. Glass type, as a key component of the building envelope, was shown to exert a direct impact on thermodynamic performance. The superior performance of Low-E glass in reducing overall energy demand underscores its energy-saving potential, while the reduced air conditioning consumption observed for vacuum and coated glass suggests their advantages in maintaining indoor thermal comfort.

As shown in Table 6, significant variations in both total energy consumption and air conditioning energy demand were observed under different combinations of window-to-wall ratio and orientation. For south-facing windows, an increase in the window-to-wall ratio from 0 to 0.66 resulted in a rise in total energy use from 125.326 to 136.235, with a corresponding increase in air conditioning energy demand from 42.325 to 43.215. In the north-facing configuration, the lowest total energy consumption (129.365) was recorded at a window-to-wall ratio of 0.51. For west-facing windows, total energy consumption fluctuated slightly, ranging from 138.235 to 137.526 as the window ratio increased from 0 to 0.66. Similarly, in the east-facing configuration, total energy use rose from 128.236 at a window ratio of 0.16 to 135.234 at 0.66. These results demonstrate that both orientation and window-to-wall ratio exert significant influence on building energy performance. The data further reinforce the critical role of BIM-driven envelope parameter and spatial configuration optimization in green building design. Through BIM-based simulation of energy consumption under various orientations and window-to-wall ratios, accurate datasets were produced to support the integration of multiple design objectives, including total energy demand and indoor thermal comfort potential.