A parameterized design building performance automatic optimization method and system
By combining Latin hypercube sampling and Monte Carlo simulation with a building performance proxy prediction model, the problems of high computational cost and slow simulation speed in traditional building parametric design are solved, achieving efficient building performance optimization and design quality improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHONGQING GUANG MEMBRANE STRUCTURE TECH CO LTD
- Filing Date
- 2026-03-24
- Publication Date
- 2026-06-23
AI Technical Summary
Traditional building parametric design and performance optimization are computationally expensive and slow to simulate under high-dimensional design parameters, making it difficult to find the global optimal solution within the engineering timeframe and failing to meet the needs of rapid design iteration.
A building performance surrogate prediction model is adopted by combining Latin hypercube sampling and Monte Carlo simulation. The NSGA-II algorithm is used for iterative optimization to generate Pareto front solution set, and performance quantification decision analysis is performed to select the optimal design parameters.
It enables the optimization of building performance in a short time, improves computational efficiency and design quality, reduces reliance on designer experience, and the resulting design scheme has excellent performance under ideal conditions and can withstand actual disturbances.
Smart Images
Figure CN122263233A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of architectural design technology, specifically relating to a method and system for automatically optimizing building performance in parametric design. Background Technology
[0002] Driven by the concept of sustainable building, performance-based design has become a key strategy in architectural design practice. By optimizing design parameters such as building form and envelope in the early stages of scheme design, building energy consumption can be significantly reduced and indoor environmental quality improved. Traditional building parametric design and performance optimization usually rely on the designer's experience and performance simulation software (such as EnergyPlus, DesignBuilder, and OpenStudio), which faces two major technical bottlenecks: First, when the design parameters are highly dimensional (such as involving multiple variables such as envelope structure, window-to-wall ratio, and shading dimensions), traditional traversal search or orthogonal experimental methods are computationally expensive, making it difficult to find the globally optimal solution within the project timeframe; second, performance simulation software has a slow simulation speed. Directly embedding it into the optimization process would lead to excessively long optimization times, huge computational costs, and low efficiency, making it difficult to meet the needs of rapid design iteration. Summary of the Invention
[0003] The purpose of this invention is to provide a method and system for automatically optimizing building performance in parametric design, in order to solve the above-mentioned problems existing in the prior art.
[0004] To achieve the above objectives, the present invention adopts the following technical solution: Firstly, a method for automatically optimizing building performance in parametric design is provided, including: S1. Obtain the initial design parameter set of the target building, and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model; S2. Based on the parametric building information model, design variables and parameter ranges are selected, and a multidimensional hypercube design space is constructed using all selected design variables and their parameter ranges. S3. The Latin hypercube sampling method is used to sample the multidimensional hypercube design space to generate several sets of initial variable samples, and multiple random scene samples of each set of initial variable samples are generated through Monte Carlo simulation. S4. Input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance, and obtain the building performance index of each group of initial variable samples under each random scenario sample. S5. Calculate the expected performance of each group of initial variable samples using the building performance index under each random scenario sample of each group of initial variable samples; S6. With the set desired performance conditions as the optimization objective, perform iterative optimization of the initial variable samples in steps S3 to S5, and after a set number of iterations, obtain the Pareto front solution set, which contains several sets of optimization variable samples. S7. Based on the expected performance of each set of optimization variable samples in the Pareto front solution set, a performance quantification decision analysis is performed on each set of optimization variable samples in the Pareto front solution set, and the optimal optimization variable sample is selected as the target design parameter for output.
[0005] In one possible design, the method further includes: The parametric building information model is optimized and updated using the target design parameters to obtain the building information model after performance optimization.
[0006] In one possible design, the initial design parameter set includes a building form parameter set, a building envelope thermal parameter set, and a shading structure parameter set. The building form parameter set includes building orientation, window-to-wall ratio, and building length-to-width ratio. The building envelope thermal parameter set includes external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, and solar heat gain coefficient. The shading structure parameter set includes shading panel projection length and shading panel transmittance. The selection of design variables and parameter ranges based on a parametric building information model includes: In response to the user's variable selection operation, any combination of building orientation, window-to-wall ratio, building length-to-width ratio, external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, solar heat gain coefficient, sunshade overhang length, and sunshade transmittance is selected as design variables, and the parameter range of each design variable is selected.
[0007] In one possible design, the random scene samples consist of human behavior pattern parameters and meteorological data that follow a normal or discrete probability distribution.
[0008] In one possible design, the building performance proxy prediction model is obtained by training the XGBoost model on a training set. The training set contains several training samples labeled with corresponding building performance indicators. The training samples include initial variable samples and random scenario samples. The building performance indicators labeled on the training samples are obtained by simulating the corresponding training samples in a pre-set building performance simulation engine.
[0009] In one possible design, the calculation of the expected performance of each set of initial variable samples using the building performance indicators under each random scenario sample includes: Calculate the average building performance index of each initial variable sample under all random scenario samples, and take the average building performance index of each initial variable sample under all random scenario samples as the expected performance of each initial variable sample. The building performance index includes energy use intensity (EUI) and effective solar illuminance (UDI), and the expected performance includes expected energy use intensity and expected effective solar illuminance.
[0010] In one possible design, the desired performance conditions include the minimum desired energy use intensity and / or the maximum desired effective solar irradiance among the initial variable samples of the same generation. When iterating and optimizing the initial variable samples, the NSGA-II algorithm is used for iterative optimization, and the number of iterations is 50.
[0011] In one possible design, based on the expected performance of each set of optimization variable samples in the Pareto front solution set, a performance quantification decision analysis is performed on each set of optimization variable samples in the Pareto front solution set, and the optimal optimization variable sample is selected as the target design parameter for output, including: Determine the maximum and minimum expected energy use intensity of all optimization variable samples in the Pareto front solution set, as well as the maximum and minimum expected effective solar irradiance of all optimization variable samples in the Pareto front solution set. The expected energy use intensity of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected energy use intensity to obtain the normalized expected energy use intensity of each group of optimization variable samples. The expected effective solar illuminance of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected effective solar illuminance to obtain the normalized expected effective solar illuminance of each group of optimization variable samples. The normalized expected energy use intensity and normalized expected effective solar irradiance of each group of optimization variables in the Pareto front solution set are weighted and summed to obtain the weighted standardized value V of each group of optimization variables. i , where i represents the index of the sample of the optimization variable; Determine the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity and the weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity, and use the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity as the first positive reference value V. a The weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity is used as the first negative reference value V. bThe weighted standardized values of the optimization variable samples corresponding to the maximum expected effective solar illuminance and the minimum expected effective solar illuminance are determined, and the weighted standardized value of the optimization variable samples corresponding to the maximum expected effective solar illuminance is used as the second positive reference value V. c The weighted standardized value of the optimization variable sample corresponding to the minimum expected effective solar illuminance is used as the second negative reference value V. d ; Using the first positive reference value V a First negative reference value V b Second positive reference value V c Second negative reference value V d and the weighted standardized value V of each group of optimization variable samples. i Calculate the ideal closeness value C for each group of optimization variable samples. i : ; The optimal variable sample with the largest ideal closeness value is output as the target design parameter.
[0012] Secondly, a parametric design automatic building performance optimization system is provided, applicable to any of the automatic building performance optimization methods described in the first aspect above, comprising a parameter acquisition unit, a variable selection unit, a sample generation unit, an index prediction unit, a performance calculation unit, an optimization summary unit, and a quantitative decision-making unit, wherein: The parameter acquisition unit is used to acquire the initial design parameter set of the target building and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model. The variable selection unit is used to select design variables and parameter ranges based on the parametric building information model, and to construct a multidimensional hypercube design space using all selected design variables and their parameter ranges. The sample generation unit is used to sample the multidimensional hypercube design space using the Latin hypercube sampling method, generate several sets of initial variable samples, and generate multiple random scene samples for each set of initial variable samples through Monte Carlo simulation. The indicator prediction unit is used to input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance and obtain the building performance indicators of each group of initial variable samples under each random scenario sample. The performance calculation unit is used to calculate the expected performance of each set of initial variable samples using the building performance index of each set of initial variable samples under each random scenario sample. The optimization summarization unit is used to summarize the results of iterative optimization using a set of algebraic initial variable samples with the set expected performance conditions as the optimization objective, forming a Pareto front solution set. The Pareto front solution set contains several sets of optimization variable samples. The quantification decision unit is used to perform performance quantification decision analysis on each set of optimization variable samples in the Pareto front solution set based on the expected performance of each set of optimization variable samples, and select the optimal optimization variable sample as the target design parameter for output.
[0013] Thirdly, a parametric design automatic building performance optimization system is provided, including: Memory, used to store instructions; A processor is configured to read instructions stored in the memory and execute, according to the instructions, any one of the parameterized design automatic optimization methods described in the first aspect above.
[0014] Fourthly, a computer-readable storage medium is provided, on which instructions are stored, which, when executed on a computer, cause the computer to perform any one of the parametric design automatic optimization methods described in the first aspect. Simultaneously, a computer program product is also provided, which, when executed on a computer, performs any one of the parametric design automatic optimization methods described in the first aspect.
[0015] Beneficial effects: This invention selects and samples initial building design parameters, combines Monte Carlo simulation with a performance target optimization framework, quantifies the impact of parameter variables on building performance, and ensures that the optimized design not only performs well in ideal conditions but also withstands random disturbances in actual operation, exhibiting higher reliability and robustness. By introducing a machine learning proxy model to replace traditional physical simulation for performance prediction, the optimization search can be completed in a short time, greatly improving the computational efficiency in the optimization iteration process. It achieves fully automated processing of parametric modeling, performance simulation, intelligent optimization, and result selection, reducing reliance on designer experience and improving the quality of building design. Attached Figure Description
[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 This is a flowchart illustrating the method in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the system configuration in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the system configuration in Embodiment 3 of the present invention. Detailed Implementation
[0018] It should be noted that the descriptions of these embodiments are intended to aid in understanding the invention and do not constitute a limitation thereof. The specific structural and functional details disclosed herein are merely for describing exemplary embodiments of the invention. However, the invention may be embodied in many alternative forms and should not be construed as being limited to the embodiments described herein.
[0019] It should be understood that, unless otherwise explicitly specified and limited, the corresponding terms should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments according to the specific circumstances.
[0020] Specific details are provided in the following description to provide a complete understanding of the exemplary embodiments. However, those skilled in the art will understand that the exemplary embodiments can be implemented without these specific details. For example, apparatus may be shown in block diagrams to avoid obscuring the examples with unnecessary details. In other embodiments, well-known processes, structures, and techniques may be omitted with non-essential details to avoid obscuring the embodiments.
[0021] Example 1: This embodiment provides an automatic optimization method for building performance in parametric design, which can be applied to corresponding servers, such as... Figure 1 As shown, the method includes the following steps: S1. Obtain the initial design parameter set of the target building, and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model.
[0022] In practice, the server first obtains the initial design parameter set of the target building. This initial design parameter set includes a building form parameter set, a building envelope thermal parameter set, and a shading structure parameter set. The building form parameter set includes building orientation, window-to-wall ratio, and building length-to-width ratio. The building envelope thermal parameter set includes the heat transfer coefficient of the exterior walls, the heat transfer coefficient of the roof, the heat transfer coefficient of the exterior windows, and the solar heat gain coefficient. The shading structure parameter set includes the shading panel projection length and the shading panel transmittance. Then, based on BIM (Building Information Modeling) parametric tools, BIM parametric modeling is performed using the initial design parameter set to obtain a parametric building information model.
[0023] S2. Based on the parametric building information model, design variables and parameter ranges are selected, and a multidimensional hypercube design space is constructed using all selected design variables and their parameter ranges.
[0024] In practice, the server can respond to the user's variable selection operation by choosing any combination of building orientation, window-to-wall ratio, building length-to-width ratio, external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, solar heat gain coefficient, shading cantilever length, and shading transmittance as design variables. The server can then select parameter ranges for each design variable. For example, the building orientation range could be 0° to 360°, the window-to-wall ratio range could be 0.2 to 0.8, the building length-to-width ratio range could be 1:1 to 3:1, and the external wall heat transfer coefficient range could be 0.15 to 0.8 W / (m²). 2 K), the roof heat transfer coefficient ranges from 0.1 to 0.5 W / (m²). 2 • K), the heat transfer coefficient of the external window ranges from 1.0 to 3.0 W / (m²). 2 K), the solar heat gain coefficient ranges from 0.2 to 0.7, the shading panel cantilever length ranges from 0.2 to 1.5 m, and the shading panel transmittance ranges from 0.1 to 0.9. Then, a multidimensional hypercube design space (multidimensional vector) is constructed using all selected design variables and their parameter ranges.
[0025] S3. The Latin hypercube sampling method is used to sample the multidimensional hypercube design space to generate several sets of initial variable samples, and multiple random scene samples of each set of initial variable samples are generated through Monte Carlo simulation.
[0026] In practice, the server can use the Latin hypercube sampling method to sample the multidimensional hypercube design space, generating several sets of initial variable samples. Each set of initial variable samples includes corresponding combinations of building orientation, window-to-wall ratio, building length-to-width ratio, external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, solar heat gain coefficient, shading cantilever length, and shading transmittance, along with their respective sampled values. Then, Monte Carlo simulation is used to generate multiple random scene samples for each set of initial variable samples. These random scene samples consist of personnel behavior pattern parameters and meteorological data that follow a normal or discrete probability distribution.
[0027] S4. Input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance, and obtain the building performance index of each group of initial variable samples under each random scenario sample.
[0028] In practice, the server pre-builds an XGBoost model and then trains it using a pre-set training set to obtain a building performance surrogate prediction model for predicting building performance. The training set contains several training samples labeled with corresponding building performance indicators. These training samples include initial variable samples (obtained using the same sampling method as in steps S1-S3) and random scenario samples. The building performance indicators labeled in the training samples are input into a pre-set building performance simulation engine (such as EnergyPlus, DesignBuilder, or IES). <ve>The results were obtained through simulation in (etc.).
[0029] In actual building performance prediction, the server will input the initial variable samples of each group and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance, and obtain the building performance index of each initial variable sample under each random scenario sample. The building performance index may include energy use intensity (EUI) and effective solar illuminance (UDI).
[0030] S5. Calculate the expected performance of each initial variable sample using the building performance index of each initial variable sample under each random scenario sample.
[0031] In practice, the server can calculate the average building performance index of each group of initial variable samples under all random scenario samples, and use the average building performance index of each group of initial variable samples under all random scenario samples as the expected performance of each group of initial variable samples. The expected performance includes expected energy use intensity and expected effective solar illuminance.
[0032] S6. With the set desired performance conditions as the optimization objective, perform iterative optimization of the initial variable samples in steps S3 to S5, and after a set number of iterations, obtain the Pareto front solution set, which contains several sets of optimization variable samples.
[0033] In practice, the server can set the desired performance conditions as the optimization objective, and use the NSGA-II algorithm to iteratively optimize the initial variable samples in steps S3 to S5. The desired performance conditions include minimizing the desired energy usage intensity and / or maximizing the desired effective solar illuminance among the initial variable samples of the same generation. After a set number of iterations, such as 50 generations, the optimized initial variable samples are obtained as the optimization variable samples, and the Pareto front solution set is formed based on the sum of these optimization variable samples.
[0034] S7. Based on the expected performance of each set of optimization variable samples in the Pareto front solution set, a performance quantification decision analysis is performed on each set of optimization variable samples in the Pareto front solution set, and the optimal optimization variable sample is selected as the target design parameter for output.
[0035] In practice, the server can determine the maximum and minimum expected energy use intensity of all optimization variable samples in the Pareto front solution set, as well as the maximum and minimum expected effective solar irradiance of all optimization variable samples in the Pareto front solution set. The expected energy use intensity of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected energy use intensity to obtain the normalized expected energy use intensity of each group of optimization variable samples. The expected effective solar illuminance of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected effective solar illuminance to obtain the normalized expected effective solar illuminance of each group of optimization variable samples. The normalized expected energy use intensity and normalized expected effective solar irradiance of each group of optimization variables in the Pareto front solution set are weighted and summed to obtain the weighted standardized value V of each group of optimization variables. i , where i represents the index of the sample of the optimization variable; Determine the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity and the weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity, and use the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity as the first positive reference value V. a The weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity is used as the first negative reference value V. b The weighted standardized values of the optimization variable samples corresponding to the maximum expected effective solar illuminance and the minimum expected effective solar illuminance are determined, and the weighted standardized value of the optimization variable samples corresponding to the maximum expected effective solar illuminance is used as the second positive reference value V. c The weighted standardized value of the optimization variable sample corresponding to the minimum expected effective solar illuminance is used as the second negative reference value V. d ; Using the first positive reference value V a First negative reference value V b Second positive reference value V c Second negative reference value V d and the weighted standardized value V of each group of optimization variable samples. i Calculate the ideal closeness value C for each group of optimization variable samples. i : ; The optimal variable sample with the largest ideal closeness value is output as the target design parameter.
[0036] Meanwhile, the server can use the final target design parameters to optimize and update the parametric building information model, resulting in a building information model with optimized building performance.
[0037] Example 2: This embodiment provides an automatic building performance optimization system for parametric design, applied to the automatic building performance optimization method in Embodiment 1, such as... Figure 2 As shown, the system includes a parameter acquisition unit, a variable selection unit, a sample generation unit, an indicator prediction unit, a performance calculation unit, an optimization and summarization unit, and a quantitative decision-making unit, wherein: The parameter acquisition unit is used to acquire the initial design parameter set of the target building and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model. The variable selection unit is used to select design variables and parameter ranges based on the parametric building information model, and to construct a multidimensional hypercube design space using all selected design variables and their parameter ranges. The sample generation unit is used to sample the multidimensional hypercube design space using the Latin hypercube sampling method, generate several sets of initial variable samples, and generate multiple random scene samples for each set of initial variable samples through Monte Carlo simulation. The indicator prediction unit is used to input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance and obtain the building performance indicators of each group of initial variable samples under each random scenario sample. The performance calculation unit is used to calculate the expected performance of each set of initial variable samples using the building performance index of each set of initial variable samples under each random scenario sample. The optimization summarization unit is used to summarize the results of iterative optimization using a set of algebraic initial variable samples with the set expected performance conditions as the optimization objective, forming a Pareto front solution set. The Pareto front solution set contains several sets of optimization variable samples. The quantification decision unit is used to perform performance quantification decision analysis on each set of optimization variable samples in the Pareto front solution set based on the expected performance of each set of optimization variable samples, and select the optimal optimization variable sample as the target design parameter for output.
[0038] This method selects and samples initial building design parameters, combines Monte Carlo simulation with a performance target optimization framework, and quantifies the impact of parameter variables on building performance. This ensures that the optimized design not only performs well under ideal conditions but also withstands random disturbances during actual operation, exhibiting higher reliability and robustness. By introducing a machine learning proxy model to replace traditional physical simulation for performance prediction, the optimization search can be completed in a short time, greatly improving computational efficiency during the optimization iteration process. It achieves fully automated processing of parametric modeling, performance simulation, intelligent optimization, and result selection, reducing reliance on designer experience and improving the quality of building design.
[0039] Example 3: This embodiment provides an automatic optimization system for building performance in parametric design, such as... Figure 3 As shown, at the hardware level, it includes: The data interface is used to establish data communication between the processor and external data terminals; Memory, used to store instructions; The processor is used to read instructions stored in the memory and execute the automatic building performance optimization method in Embodiment 1 according to the instructions.
[0040] Optionally, the system also includes an internal bus, through which the processor, memory, and data interface can be interconnected. This internal bus can be a PCIe (Peripheral Component Interconnect Eexpress) bus, which can be divided into an address bus, a data bus, a control bus, etc. The memory can include, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Flash Memory, First Input First Output (FIFO), and / or First In Last Out (FILO). The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0041] Example 4: This embodiment provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the automatic building performance optimization method of Embodiment 1. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or Memory Sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
[0042] This embodiment also provides a computer program product that, when run on a computer, executes the automatic building performance optimization method described in Embodiment 1. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.
[0043] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.< / ve>
Claims
1. A method for automatically optimizing building performance in parametric design, characterized in that, include: S1. Obtain the initial design parameter set of the target building, and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model; S2. Based on the parametric building information model, design variables and parameter ranges are selected, and a multidimensional hypercube design space is constructed using all selected design variables and their parameter ranges. S3. The Latin hypercube sampling method is used to sample the multidimensional hypercube design space to generate several sets of initial variable samples, and multiple random scene samples of each set of initial variable samples are generated through Monte Carlo simulation. S4. Input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance, and obtain the building performance index of each group of initial variable samples under each random scenario sample. S5. Calculate the expected performance of each group of initial variable samples using the building performance index under each random scenario sample of each group of initial variable samples; S6. With the set desired performance conditions as the optimization objective, perform iterative optimization of the initial variable samples in steps S3 to S5, and after a set number of iterations, obtain the Pareto front solution set, which contains several sets of optimization variable samples. S7. Based on the expected performance of each set of optimization variable samples in the Pareto front solution set, a performance quantification decision analysis is performed on each set of optimization variable samples in the Pareto front solution set, and the optimal optimization variable sample is selected as the target design parameter for output.
2. The method for automatic optimization of building performance in parametric design according to claim 1, characterized in that, The method further includes: The parametric building information model is optimized and updated using the target design parameters to obtain the building information model after performance optimization.
3. The automatic optimization method for building performance in parametric design according to claim 1, characterized in that, The initial design parameter set includes a building form parameter set, a building envelope thermal parameter set, and a shading structure parameter set. The building form parameter set includes building orientation, window-to-wall ratio, and building length-to-width ratio. The building envelope thermal parameter set includes external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, and solar heat gain coefficient. The shading structure parameter set includes shading panel projection length and shading panel transmittance. The selection of design variables and parameter ranges based on the parametric building information model includes: In response to the user's variable selection operation, any combination of building orientation, window-to-wall ratio, building length-to-width ratio, external wall heat transfer coefficient, roof heat transfer coefficient, external window heat transfer coefficient, solar heat gain coefficient, sunshade overhang length, and sunshade transmittance is selected as design variables, and the parameter range of each design variable is selected.
4. The method for automatic optimization of building performance in parametric design according to claim 1, characterized in that, The random scene samples consist of human behavior pattern parameters and meteorological data that follow a normal distribution or a discrete probability distribution.
5. The automatic optimization method for building performance in parametric design according to claim 1, characterized in that, The building performance proxy prediction model is obtained by training the XGBoost model on a training set. The training set contains several training samples labeled with corresponding building performance indicators. The training samples include initial variable samples and random scenario samples. The building performance indicators labeled on the training samples are obtained by simulating the corresponding training samples in a pre-set building performance simulation engine.
6. The method for automatic optimization of building performance in parametric design according to claim 1, characterized in that, The calculation of the expected performance of each set of initial variable samples using the building performance indicators of each set of initial variable samples under each random scenario sample includes: Calculate the average building performance index of each initial variable sample under all random scenario samples, and take the average building performance index of each initial variable sample under all random scenario samples as the expected performance of each initial variable sample. The building performance index includes energy use intensity (EUI) and effective solar illuminance (UDI), and the expected performance includes expected energy use intensity and expected effective solar illuminance.
7. The automatic optimization method for building performance in parametric design according to claim 6, characterized in that, The desired performance conditions include the minimum desired energy use intensity and / or the maximum desired effective solar irradiance among the initial variable samples of the same generation. When iteratively optimizing the initial variable samples, the NSGA-II algorithm is used for iterative optimization, and the number of iterations is 50.
8. The method for automatic optimization of building performance in parametric design according to claim 7, characterized in that, The method involves analyzing the expected performance of each set of optimization variable samples in the Pareto front solution set, performing performance quantification decision analysis on each set of optimization variable samples, and selecting the optimal optimization variable sample as the target design parameter for output. This includes: Determine the maximum and minimum expected energy use intensity of all optimization variable samples in the Pareto front solution set, as well as the maximum and minimum expected effective solar irradiance of all optimization variable samples in the Pareto front solution set. The expected energy use intensity of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected energy use intensity to obtain the normalized expected energy use intensity of each group of optimization variable samples. The expected effective solar illuminance of each group of optimization variable samples in the Pareto front solution set is normalized using the maximum and minimum expected effective solar illuminance to obtain the normalized expected effective solar illuminance of each group of optimization variable samples. The normalized expected energy use intensity and normalized expected effective solar irradiance of each group of optimization variables in the Pareto front solution set are weighted and summed to obtain the weighted standardized value V of each group of optimization variables. i , where i represents the index of the sample of the optimization variable; Determine the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity and the weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity, and use the weighted standardized value of the optimization variable sample corresponding to the minimum expected energy use intensity as the first positive reference value V. a The weighted standardized value of the optimization variable sample corresponding to the maximum expected energy use intensity is used as the first negative reference value V. b The weighted standardized values of the optimization variable samples corresponding to the maximum expected effective solar illuminance and the minimum expected effective solar illuminance are determined, and the weighted standardized value of the optimization variable samples corresponding to the maximum expected effective solar illuminance is used as the second positive reference value V. c The weighted standardized value of the optimization variable sample corresponding to the minimum expected effective solar illuminance is used as the second negative reference value V. d ; Using the first positive reference value V a First negative reference value V b Second positive reference value V c Second negative reference value V d and the weighted standardized value V of each group of optimization variable samples. i Calculate the ideal closeness value C for each group of optimization variable samples. i : ； The optimal variable sample with the largest ideal closeness value is output as the target design parameter.
9. A parametric design automatic building performance optimization system, applied to the automatic building performance optimization method according to any one of claims 1-8, characterized in that, It includes a parameter acquisition unit, a variable selection unit, a sample generation unit, an indicator prediction unit, a performance calculation unit, an optimization and summarization unit, and a quantitative decision-making unit, wherein: The parameter acquisition unit is used to acquire the initial design parameter set of the target building and use the initial design parameter set to perform BIM parametric modeling to obtain a parametric building information model. The variable selection unit is used to select design variables and parameter ranges based on the parametric building information model, and to construct a multidimensional hypercube design space using all selected design variables and their parameter ranges. The sample generation unit is used to sample the multidimensional hypercube design space using the Latin hypercube sampling method, generate several sets of initial variable samples, and generate multiple random scene samples for each set of initial variable samples through Monte Carlo simulation. The indicator prediction unit is used to input each group of initial variable samples and each random scenario sample into the pre-trained building performance proxy prediction model to predict building performance and obtain the building performance indicators of each group of initial variable samples under each random scenario sample. The performance calculation unit is used to calculate the expected performance of each set of initial variable samples using the building performance index of each set of initial variable samples under each random scenario sample. The optimization summarization unit is used to summarize the results of iterative optimization using a set of algebraic initial variable samples with the set expected performance conditions as the optimization objective, forming a Pareto front solution set. The Pareto front solution set contains several sets of optimization variable samples. The quantification decision unit is used to perform performance quantification decision analysis on each set of optimization variable samples in the Pareto front solution set based on the expected performance of each set of optimization variable samples, and select the optimal optimization variable sample as the target design parameter for output.
10. A parametric design automatic optimization system for building performance, characterized in that, include: Memory, used to store instructions; A processor is configured to read instructions stored in the memory and execute the automatic building performance optimization method according to any one of claims 1-8.