Multi-objective Optimization Design Method and Platform for Prefabricated Steel Structure Ultra-low Energy Consumption Buildings

Abstract

The present invention relates to a multi-objective optimization design method and platform for prefabricated steel structure ultra-low energy consumption buildings. Considering the two aspects of cost and energy consumption of prefabricated steel structure ultra-low energy consumption buildings, the specific content includes the following parts: According to the structural characteristics of prefabricated steel structures, parametric design research and automatic modeling of BIM dynamic models are carried out; Considering the influence of building envelope structure material parameters on building energy consumption, a variety of artificial intelligence algorithms are used to complete the prediction of building energy consumption. Through comparison, an efficient and accurate building energy consumption prediction method is obtained and variable importance analysis is carried out; Considering the bearing capacity constraint conditions of the structure, a multi-objective optimization algorithm is used for multi-objective optimization design of prefabricated steel structure ultra-low energy consumption buildings. At the same time, a multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings is constructed, which organically combines structural stability checking, embedded machine learning models and multi-objective optimization design, making the entire actual process more automated and intelligent.

Description

Technical Field

[0001] The present invention belongs to the field of building construction, and particularly relates to a multi-objective optimization design method and platform for prefabricated steel structure ultra-low energy consumption buildings. Background Technique

[0002] Since the reform and opening up, China's construction industry has played an important role in promoting social and economic development, urban and rural construction, and improvement of the human settlement environment. Compared with the growing needs of the people for a better life, there is still huge room for development in the construction industry in terms of scientific and technological innovation, improving efficiency, enhancing quality, reducing pollution and emissions. The preliminary design stage of prefabricated steel structure ultra-low energy consumption buildings is an important stage determining the building performance. However, there is no effective solution for determining the best structural parameters when considering multiple design objectives simultaneously. Parametric design has not yet formed a perfect design system in prefabricated steel structure ultra-low energy consumption buildings.

[0003] Building operation energy consumption accounts for more than 30% of the total energy consumption in China. In order to promote China to achieve the carbon peak and carbon neutrality goals planned in the "14th Five-Year Plan" as scheduled, in addition to the safety and stability check of the building structure design, the building energy consumption corresponding to the structure should also be considered. However, the higher the quality of building materials, the higher the cost and the lower the energy consumption. Therefore, there is often a conflict between the design objectives of reducing building costs and reducing building energy consumption. There is no effective research to solve this multi-objective optimization design problem, resulting in the design scheme being difficult to meet the market's expectations of both energy conservation and economy. And some relevant scholars at home and abroad have conducted preliminary research on the multi-objective optimization design of building structures. Most of them obtain the optimal solution set of the multi-objective function based on gradient optimization or by changing the weighting coefficient, and apply the meta-heuristic algorithm to the structural optimization design. However, these studies have a high computational cost and take a long time, and there is still a lack of research on the cost and energy consumption of prefabricated steel structure ultra-low energy consumption buildings.

[0004] In view of this, starting from the actual engineering situation, the present invention aims to invent a multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings that can solve the above-mentioned shortcomings in the prior art for new buildings. Summary of the Invention

[0005] The purpose of the present invention is to provide a multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings.

[0006] To achieve the above purpose, the technical solution of the present invention is:

[0007] A multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings, considering two aspects of the cost and energy consumption of prefabricated steel structure ultra-low energy consumption buildings, is characterized in that the specific process of this design method is as follows:

[0008] 1) Parametric modeling:

[0009] Due to the modular and standardized characteristics of integrating room module units in prefabricated steel structure ultra-low energy consumption buildings, different building total lengths and different building total widths have different household layout methods. Fully considering the minimization of cantilevers, the principle of parametric layout, and the structural assembly method at the construction site, the household layout that automatically matches the current building total length and building total width is achieved;

[0010] By adjusting different building total lengths and building total widths, a prefabricated steel structure building that meets specific goals is automatically generated. By adjusting the number of building floors, the household layout of the entire prefabricated steel structure can be completed, and then the material properties are assigned to the beams and columns to obtain the BIM model of the steel frame;

[0011] The BIM model of the steel frame is calculated for structural stability according to the specifications, and the total steel consumption of the steel structure frame in the BIM model of the steel frame is counted, and the steel frame cost in the total cost calculation is obtained according to the steel unit price;

[0012] 2) Analysis of influencing factors of building energy consumption:

[0013] Using building energy consumption simulation software, analyze the building energy consumption changes of the performance parameters of the building envelope structure materials that affect building energy consumption one by one, and determine the types and respective value ranges of the key performance parameters of the building envelope structure materials that affect building energy consumption;

[0014] 3) Building energy consumption prediction:

[0015] Build a benchmark building energy consumption model, select the weather file in the energy consumption simulation software according to the engineering region, use the key performance parameters of the building envelope structure materials as independent variables, and the corresponding energy consumption values simulated by the energy consumption simulation software as dependent variables, and vary within the respective value ranges of the key performance parameters of the building envelope structure materials to obtain a building energy consumption database corresponding to the key performance parameters of the building envelope structure materials and energy consumption values;

[0016] Build an artificial intelligence algorithm database, including at least a number of intelligent algorithms such as XGBoost algorithm (XGBoost), Gradient Boosting algorithm (GBR), Random Forest algorithm (RFR), ExtraTrees algorithm (ETR), Gaussian Process algorithm (GPR), and K-Nearest Neighbor Regression algorithm (KNR);

[0017] Using the building energy consumption database, with the key performance parameters of the building envelope structure materials as inputs and the energy consumption values as outputs, train various intelligent algorithms in the artificial intelligence algorithm library to find the current optimal intelligent algorithm as the building energy consumption prediction model;

[0018] Find the contribution values of each key performance parameter of the building envelope structure materials in the building energy consumption prediction model, and determine the two key performance parameters of the building envelope structure materials that have the greatest impact on building energy consumption;

[0019] 4) Multi-objective optimization

[0020] Construct a BIM parametric library with the total building length, total building width, total number of building floors, column size, beam size, and performance parameters of key envelope structure materials. Based on the data in the BIM parametric library, use the building cost per unit area Cost(x) and building energy consumption per unit area Energy(x) as the objective functions of optimization, and construct a multi-objective optimization design calculation model according to the following formula;

[0021] Cost(x) = TC / (L * W * N)

[0022] Energy(x) = E / (L * W * N)

[0023] Where, TC - economic cost of residential building; L - total building length; W - total building width; N - total number of building floors; E - total building energy consumption obtained from the building energy consumption prediction model in step 3);

[0024] Economic cost of residential building TC:

[0025]

[0026] Where: C b — cost of foundation wall; C win — cost of window; C r — cost of roof insulation; C w — cost of external wall insulation; C d — cost of floor insulation; r - discount rate; y i , z i — replacement life; C s — system cost; C 钢 — cost of steel frame, obtained by multiplying the steel consumption by the steel unit price obtained in the parametric modeling in step 1);

[0027] The constraint conditions for multi-objective optimization are the thermal coefficient limit standard constraints of the performance parameters of key envelope structure materials and the strength, stiffness, and stability constraints of the steel frame structure;

[0028] Use the multi-objective optimization algorithm to obtain the optimal solution, put the optimal solution into the BIM model of the steel frame in the parametric modeling process, obtain the final BIM model, and realize the optimized design of prefabricated steel structure ultra-low energy consumption buildings.

[0029] The performance parameters of the key envelope structure materials include the thickness of the external wall insulation layer, the thickness of the roof insulation layer, the thickness of the floor insulation layer, the east-west window wall ratio, the west-east window wall ratio, the south-north window wall ratio, the north-south window wall ratio, and the type of external window;

[0030] The strength, stiffness and stability constraints of the steel frame structure mean meeting the following conditions:

[0031]

[0032]

[0033] v b ≤[v b =I b / 400

[0034] λ cx =l cx / i cx ≤[λ]; λ cy =l cy / i cy ≤[λ]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] 1×10 5 ≤I b ≤5×10 9

[0041] 1×10 6 ≤I c1 ,I c2 ≤5×10 9

[0042] Where: M bx — Design value of the bending moment about the x-axis at the same section; W bx — Net section modulus about the x-axis; γ x — Section plastic development coefficient about the main axis x; f — Design value of the flexural and compressive strength of steel; N c — Design value of the axial force of the column (N); A c — Cross-sectional area of the column (mm 2 ); W cx — Section modulus of the column (mm 3 ); l cx , l cy — Calculated lengths of the column in the x-axis and y-axis; β mx , β tx — Equivalent bending moment coefficients for in-plane and out-of-plane stability checks; — The stability coefficients of axially compressed members in and out of the plane of bending moment; Δ T — The imaginary displacement at the top of the structure; x and y are the x-axis and y-axis during the BIM model modeling of the steel frame.

[0043] Using a variety of artificial intelligence algorithms to establish a building energy consumption prediction model, and using the mean absolute error (MAE) and the coefficient of determination R 2 as the accuracy performance evaluation indicators of the machine learning model for evaluation.

[0044] The present invention also protects a multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings, which is characterized in that the design platform includes:

[0045] A parametric modeling module, which is used to determine the layout method of the housing type of the prefabricated steel structure according to the characteristics of the prefabricated steel structure ultra-low energy consumption building, and establish a BIM model of the steel frame according to the material properties of the beam and column;

[0046] A building energy consumption influencing factor analysis module, which is used to determine the types and respective value ranges of the key envelope structure material performance parameters that affect building energy consumption by using building energy consumption simulation software;

[0047] An artificial intelligence algorithm database, which contains various intelligent algorithms;

[0048] A building energy consumption prediction module, which constructs a building energy consumption database according to the types and respective value ranges of the key envelope structure material performance parameters determined by the building energy consumption influencing factor analysis module, and at the same time uses the building energy consumption database to train and evaluate various intelligent algorithms in the artificial intelligence algorithm database, and uses the intelligent algorithm with the optimal performance as the building energy consumption prediction model;

[0049] A multi-objective optimization module, which forms a BIM parameterization library with the total building length, total building width, total number of building floors, column size, beam size, and key envelope structure material performance parameters, and based on the data in the BIM parameterization library, uses the unit area building cost Cost(x) and the unit area building energy consumption Energy(x) as the optimization objective functions for multi-objective optimization calculation to determine the optimal solution set; the steel frame cost calculated from the steel consumption in the BIM model of the steel frame is considered in the unit area building cost Cost(x), and the constraints of the multi-objective optimization are the thermal coefficient limit standard constraints of the key envelope structure material performance parameters and the strength, stiffness, and stability constraints of the steel frame structure;

[0050] A scheme generation module, which applies the scheme in the optimal solution set obtained by the multi-objective optimization module to the BIM model of the steel frame in the parametric modeling module to obtain the final BIM model.

[0051] The design platform further includes a structural stability calculation module for calculating the stability of the BIM model of the steel frame.

[0052] The parametric modeling module is implemented through the Dynamo visual programming module, and the multi-objective optimization module is implemented through the programmable module Python Script in Dynamo visual programming.

[0053] The present invention protects a computer-readable storage medium with a computer program stored thereon, characterized in that when the program is executed by a processor, the steps of the multi-objective optimization design method for the prefabricated steel structure ultra-low energy consumption building described above can be realized.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] ① Fully consider the influence of the basic information of the building on the building structure layout, that is, the total length, total width and number of floors of the building. Through parametric modeling, a dynamic and logically accurate BIM model of the steel frame is automatically generated. By inputting the total length, total width and number of floors of the building, a logically accurate BIM model of the steel frame that meets specific requirements can be automatically output, avoiding repeated modeling work, and calculating the total steel consumption of the steel structure frame. At the same time, a multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings is constructed, which organically combines structural stability checking, embedding machine learning models and multi-objective optimization design, making the entire actual process more automated and intelligent.

[0056] ② Apply a variety of artificial intelligence algorithms to establish an energy consumption prediction model, combined with building energy consumption simulation software, such as Energyplus, DOE-2, Equest, Visusl DOE, etc. Combining with actual projects, through the comparison of a variety of artificial intelligence algorithms, an efficient and reliable building energy consumption prediction method is obtained. By inputting the parameters of the building's envelope structure, the total energy consumption of the building can be output, and further calculate the contribution values of each feature value in the machine learning model, and quantitatively analyze the importance of each feature to the building energy consumption calculation, which is convenient for later design adjustment.

[0057] ③ Use multi-objective optimization algorithms, such as non-dominated sorting genetic algorithm (NSGA-II), multi-objective particle swarm optimization algorithm (PSO), etc., for the selected artificial intelligence algorithm model. For prefabricated steel structure ultra-low energy consumption buildings, combined with the lowest cost and energy consumption, considering the bearing capacity requirements of the structure, a multi-objective optimization design method for cost and energy consumption is realized on the premise of meeting structural safety.

[0058] ④ Statistically analyze the value rules of the design parameters in the optimal solution set scheme after multi-objective optimization, and compare them with the variable importance analysis statistically obtained by the artificial intelligence algorithm, which greatly increases the accuracy of the multi-objective optimization algorithm and the artificial intelligence algorithm calculation, and ensures the feasibility and reliability of the design method. Brief Description of the Drawings

[0059] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention, and thus should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0060] Figure 1 It is a schematic diagram of the parametric process of the coupled-column modular steel frame structure of the present invention.

[0061] Figure 2 It is a schematic diagram of the database model of the present invention.

[0062] Figure 3 It is a flowchart of the machine learning modeling of the present invention.

[0063] Figure 4 It is an analysis of the importance of predictive variables during the training process of the present invention.

[0064] Figure 5 It is a flowchart of the optimization process of the present invention.

[0065] Figure 6 It is a schematic diagram of the Pareto solution set frontier distribution of the present invention.

[0066] Figure 7 It is a frequency distribution diagram of the values of each parameter of the present invention in the Pareto optimal solution. Detailed Embodiments

[0067] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, taking the typical prefabricated steel structure ultra-low energy consumption building - the coupled-column modular prefabricated steel structure ultra-low energy consumption building as an example, combined with the drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described.

[0068] The multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings of the present invention considers two aspects: the cost and energy consumption of prefabricated steel structure ultra-low energy consumption buildings. The specific content includes the following parts: According to the structural characteristics of prefabricated steel structures, parametric design research and automatic modeling of BIM dynamic models are carried out; considering the influence of building envelope structure material parameters on building energy consumption, a variety of artificial intelligence algorithms are used to complete the prediction of building energy consumption, and an efficient and accurate building energy consumption prediction method is obtained through comparison and variable importance analysis; considering the bearing capacity constraint conditions of the structure, a multi-objective optimization algorithm is used for multi-objective optimization design of prefabricated steel structure ultra-low energy consumption buildings.

[0069] 1) Based on the characteristics of prefabricated steel structure ultra-low energy consumption buildings, establish a visualization model and automatically generate a dynamic and logically accurate BIM model. Statistically calculate the total steel consumption of the steel structure frame in the BIM model of the steel frame, and multiply it by the steel unit price in the market to provide the steel frame cost for the cost calculation model in the multi-objective optimization design.

[0070] 2) Based on building energy consumption simulation software, analyze the performance parameters of the exterior envelope structure materials that affect building energy consumption one by one, and statistically calculate the change range of each parameter according to relevant technical specifications.

[0071] 3) Conduct predictive analysis of building energy consumption through various artificial intelligence algorithms, aiming to find an efficient and reliable building energy consumption prediction model.

[0072] 4) Utilize the steel frame cost obtained from parametric modeling and the building energy consumption prediction model, and combine with the market price of building materials to construct a multi-objective optimization calculation model, and use intelligent algorithms to obtain the optimal solution set that meets the conditions.

[0073] The specific process of the multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings is as follows:

[0074] 1) Parametric modeling:

[0075] Based on the characteristics of prefabricated steel structure ultra-low energy consumption buildings, establish a visualization model and automatically generate a dynamic and logically accurate BIM model.

[0076] Due to the modular and standardized characteristics of the integrated room module units, different building lengths and different building widths have different household layout methods. Fully consider the principles of minimizing cantilevers, parametric layout, and the structural assembly method at the construction site, and automatically match the household layout suitable for the current building length and building width;

[0077] Generate prefabricated steel structure buildings that meet specific goals by adjusting different building lengths and building widths. By adjusting the number of building floors, the household layout of the entire prefabricated steel structure can be completed. Then, assign material properties to the beams and columns to obtain the BIM model of the steel frame, and immediately perform structural stability calculations according to the specifications;

[0078] Perform structural stability calculations on the BIM model of the steel frame according to the specifications, and statistically calculate the total steel consumption of the steel structure frame in the BIM model of the steel frame. Investigate the steel unit price in the market, multiply the two, and obtain the steel frame cost in the total cost calculation according to the steel unit price.

[0079] 2) Analysis of factors affecting building energy consumption:

[0080] Based on building energy consumption simulation software, analyze the performance parameters of the envelope structure materials that affect building energy consumption one by one. Determine the types and respective value ranges of the key envelope structure material performance parameters that affect building energy consumption by whether there is an obvious change in building energy consumption.

[0081] 3) Building energy consumption prediction:

[0082] Fully consider the optimized design of the structural dimensions for building energy consumption control. Combine the performance parameters of the key envelope structure materials to construct a benchmark building energy consumption model. Take the performance parameter ranges of the main envelope structure materials that comply with the "Energy Conservation Standard for Residential Buildings in Hebei Province" and have an impact on building energy consumption as the value ranges of each material parameter in the database, and establish a building energy consumption database for actual engineering cases.

[0083] Construct a benchmark building energy consumption model. Select the weather file in the energy consumption simulation software according to the engineering region. Use the performance parameters of the key envelope structure materials as independent variables and the corresponding energy consumption values simulated by the energy consumption simulation software as dependent variables, and vary within their respective value ranges to obtain a large amount of data corresponding to the performance parameters of the key envelope structure materials and energy consumption values, and thus construct a building energy consumption database;

[0084] Construct an artificial intelligence algorithm library, including XGBoost algorithm (XGBoost), Gradient Boosting algorithm (GBR), Random Forest algorithm (RFR), ExtraTrees algorithm (ETR), Gaussian Process algorithm (GPR), and K-Nearest Neighbor Regression algorithm (KNR), etc.;

[0085] Use the constructed building energy consumption database to train various algorithms in the artificial intelligence algorithm library, find the current optimal intelligent algorithm as the building energy consumption prediction model, find the contribution values of each feature (performance parameters of the key envelope structure materials) based on the optimal intelligent algorithm, and determine the two variables that have the greatest impact on building energy consumption;

[0086] Apply multiple artificial intelligence algorithms to establish a building energy consumption prediction model. Through multiple-dimensional evaluation criteria, compare and analyze to obtain an efficient and accurate building energy consumption simulation prediction model suitable for the current reality. Further, statistically analyze the contribution values of each feature during the model training process, and quantitatively analyze the importance of each feature to the calculation of building energy consumption. The binary pkl file and the dat file storing data derived from the building energy consumption prediction model will be used for the building energy consumption calculation model in the multi-objective optimization process.

[0087] 4) Multi-objective optimization design:

[0088] The total building length, total building width, number of building floors, column size, beam size, thickness of the external wall insulation layer, thickness of the roof insulation layer, thickness of the ground insulation layer, east-facing window-wall ratio, west-facing window-wall ratio, south-facing window-wall ratio, and north-facing window-wall ratio are used to form a BIM parametric library. Based on the data in the BIM parametric library, the unit area building cost and unit area building energy consumption are used as the optimization objective functions, and a multi-objective optimization calculation model is constructed according to the following formula

[0089] Cost(x) = TC / (L*W*N)

[0090] Energy(x) = E / (L*W*N)

[0091] The multi-objective optimization is carried out using the multi-objective optimization algorithm NSGA-II to obtain the optimal solution, and the optimal solution is put into the BIM model of the steel frame for parametric modeling to obtain the final BIM model.

[0092] Using the steel consumption and building energy consumption prediction models of the steel frame obtained by parametric modeling, the unit prices of various external wall enclosure structure materials are investigated to obtain the cost and energy consumption of the prefabricated steel structure ultra-low energy consumption building. Using the unit area building cost and unit area building energy consumption as the optimization objective functions ensures the objectivity of the optimization. The multi-objective design of the prefabricated steel structure ultra-low energy consumption building is carried out using the multi-objective optimization algorithm. The constraint conditions of the performance parameters of the external wall enclosure structure materials have been considered in the building energy consumption prediction model, and the steel frame structure is subject to the constraint condition of structural bearing capacity requirements, and the optimal solution set of the minimum building cost and minimum building energy consumption that meet the structural bearing capacity and the requirements of the "Energy Conservation Standard for Residential Buildings in Hebei Province" is obtained. Subsequently, the distribution frequencies of the values of each design parameter in the obtained optimal solution set are statistically analyzed to study the value-taking rules of the design parameters in the optimal solution set, providing a basis for adjusting the optimized design parameters.

[0093] The conclusion obtained by the multi-objective optimization through the adjustment parameters determined by the distribution frequency is consistent with the variable importance determined by the artificial intelligence algorithm, which proves the correctness of the algorithm.

[0094] The final optimization scheme can automatically generate the corresponding BIM model through the established parametric modeling. The multi-objective optimization module will be encapsulated in the programmable module in the parametric modeling process. By selecting the scheme in the Pareto solution set, the corresponding BIM model can be directly generated, and finally, the design scheme of the coupled-column modular steel structure ultra-low energy consumption building that meets the actual project cost and energy consumption is completed.

[0095] The multi-objective optimization design platform for the prefabricated steel structure ultra-low energy consumption building provided by the present invention includes:

[0096] A parametric modeling module, which is used to determine the layout mode of housing types of prefabricated steel structures according to the characteristics of prefabricated steel structure ultra-low energy consumption buildings, and establish a BIM model of the steel frame based on the material properties of beams and columns;

[0097] A building energy consumption influencing factor analysis module, which is used to determine the types and respective value ranges of the performance parameters of key envelope structure materials that affect building energy consumption by using building energy consumption simulation software;

[0098] An artificial intelligence algorithm database, which contains various intelligent algorithms;

[0099] A building energy consumption prediction module, which constructs a building energy consumption database based on the types and respective value ranges of the performance parameters of key envelope structure materials determined by the building energy consumption influencing factor analysis module, and at the same time uses the building energy consumption database to train and evaluate various intelligent algorithms in the artificial intelligence algorithm database, and uses the intelligent algorithm with the best performance as the building energy consumption prediction model;

[0100] A multi-objective optimization module, which forms a BIM parameterization library with the total length of the building, the total width of the building, the total number of floors of the building, the column size, the beam size, and the performance parameters of key envelope structure materials. Based on the data in the BIM parameterization library, the unit area building cost Cost(x) and the unit area building energy consumption Energy(x) are used as the optimization objective functions to perform multi-objective optimization calculations to determine the optimal solution set; the steel frame cost calculated from the steel consumption in the steel frame BIM model is considered in the unit area building cost Cost(x). The constraint conditions for multi-objective optimization are the thermal coefficient limit standard constraints of the performance parameters of key envelope structure materials and the strength, stiffness, and stability constraints of the steel frame structure;

[0101] A scheme generation module, which applies the scheme in the optimal solution set obtained by the multi-objective optimization module to the BIM model of the steel frame in the parametric modeling module to obtain the final BIM model.

[0102] This design platform organically combines multi-objective optimization design, structural stability calculation, and artificial intelligence algorithms of machine learning, making the optimization design of prefabricated steel structure ultra-low energy consumption buildings faster and more convenient.

[0103] Embodiment

[0104] 1. Parametric modeling:

[0105] The parametric modeling of the coupled-wall modular steel frame structure is realized through the Dynamo visual programming module. The range of the total building length is set to [9.9, 13.2] m, the range of the total building width is [9.9, 10.8] m, and the range of the number of building floors is [1, 6] floors, meeting the requirements of the floor area of ordinary residences. Square steel tube columns are used for the columns, and hot-rolled H-shaped steel is selected for the beams. Visual modeling is achieved through Dynamo visual programming. According to the information such as the total building length, total building width, and number of floors, as well as conditions such as the size of the integrated room module unit, the size range of the coupled-wall unit, and the shape characteristics of the integrated floor unit, a coupled-wall modular building that meets specific goals is automatically generated.

[0106] First, determine the generation sequence of the modular building according to the total building length, total building width, and total number of building floors, and adjust the block module in the Dynamo visual programming through the values of the total building length, total building width, and total number of building floors. Under the constraints such as the standardized size, layout method, and cantilever minimization of the integrated room module unit, the integrated room module unit and the coupled-wall unit are established in sequence. The position of the integrated room module unit depends on the position of the center point of the integrated room module and the size of the integrated room module unit. The sizes of the integrated room module units are different under different conditions of the total building length, total building width, and number of building floors. Under the position of each integrated room module unit and the condition that the building meets the requirements of the floor area of ordinary residences, the size of the coupled-wall unit with easy adjustment can be determined, and finally, the design scheme of the entire coupled-wall modular building is completed.

[0107] Taking a two-story building as an example, the specific modeling process is as Figure 1 shown, Figure 1 In the first and second figures in [reference], the black dots are the center points of the modules, the small squares are the sizes of the integrated room module units, the third figure is the roof of the integrated room module unit formed by translating the bottom layer by the floor height, the fourth figure is the cuboid formed by connecting the roof and the ground as the integrated room module unit, the fifth figure pointed by the arrow is the coupled-wall unit formed by connecting the blank spaces between two adjacent integrated room module units. The size of the coupled-wall unit is adjustable. After arranging the first floor and then translating to form the second floor as shown in the sixth figure, and then assigning material properties to the beams and columns to form the BIM model of the steel frame as shown in the last figure.

[0108] Through Revit for visual analysis, a dynamic BIM model of the steel frame is generated, and then the structural stability calculation is carried out. The total steel consumption of the steel structure frame is counted, and the market steel unit price is investigated to be 5450 yuan / t. The two are multiplied to obtain the steel frame cost in the total cost calculation, which is used for the cost calculation of multi-objective optimization in the later stage.

[0109] 2. Analysis of influencing factors of building energy consumption:

[0110] Investigate the performance factors of the envelope structure materials that affect building energy consumption. Combine with the building energy consumption simulation software Energyplus for verification. For the performance parameters of the envelope structure materials of the coupled-column modular steel frame structure houses, determine the key envelope structure material performance parameters, including: the thickness of the external wall insulation layer, the thickness of the roof insulation layer, the thickness of the ground insulation layer, the east-west window-wall ratio, the west-east window-wall ratio, the south-north window-wall ratio, the north-south window-wall ratio, and the heat transfer coefficient of the external window. The parameter variation ranges of these 8 parameters all meet the "Energy Conservation Standard for Residential Buildings in Hebei Province", and the variation ranges are as follows:

[0111]

[0112] 3. Energy consumption simulation prediction:

[0113] The building energy consumption model takes a certain coupled-column modular steel frame structure in Xingtai City as the object. The benchmark building energy consumption model is as Figure 2 shown. The construction of the specific benchmark building energy consumption model is a well-known technology. Using the parametric variation range that meets the "Energy Conservation Standard for Residential Buildings in Hebei Province", combine with the calculation data of the energy consumption simulation software Energyplus to construct a building energy consumption database.

[0114] Apply a variety of machine learning algorithms for building energy consumption prediction. The specific machine learning algorithms include 6 machine learning algorithms: XGBoost algorithm (XGBoost), Gradient Boosting algorithm (GBR), Random Forest algorithm (RFR), ExtraTrees algorithm (ETR), Gaussian Process algorithm (GPR), and K-Nearest Neighbor Regression algorithm (KNR). The specific machine learning modeling process is as Figure 3 shown. The building energy consumption database contains 1120 groups of data sets. To ensure the accuracy and robustness of model training, this study carried out feature engineering processing such as feature normalization and heat transfer coefficient value selection of the external window type on the basis of retaining the characteristics of the original data set. Then, the database was divided into a training set and a test set according to a ratio of 7:3. Then, the training set was input into the machine learning algorithm for building energy consumption model modeling. During this period, the K-fold (K = 5) cross-validation and grid search methods were used for hyperparameter tuning of the algorithm. The objective function of the tuning is the mean squared error (MSE) of the data set. After the model training is completed, the mean absolute error (MAE) and the coefficient of determination R 2 are used as the accuracy performance evaluation indicators of the machine learning model. By comparing the evaluation indicators (mean absolute error MAE and coefficient of determination R 2 ) of the models established by the corresponding 6 machine learning algorithms applied to the training set and the test set, except for the K-Nearest Neighbor Regression algorithm, the other 5 algorithms all show good prediction ability. Among them, the XGBoost model and the Gaussian Process model have better discreteness (lower MAE). Between the two, the XGBoost model has better prediction accuracy (higher R2 ) In this embodiment, the XGBoost model is finally selected as the building energy consumption prediction model.

[0115] Then, the contribution values of each feature during the training process of the XGBoost model are statistically analyzed, and the importance of each feature for building energy consumption calculation is quantitatively analyzed. For example, Figure 4 As shown, the features in the figure are respectively: exterior wall insulation layer thickness EWI, roof insulation layer thickness RI, ground insulation layer thickness GI, east-facing window-wall ratio EWR, west-facing window-wall ratio WWR, south-facing window-wall ratio SWR, north-facing window-wall ratio NWR, and exterior window heat transfer coefficient HTC. It can be seen that the two variables that have the greatest impact on building energy consumption are the exterior wall insulation layer thickness and the roof insulation layer thickness, which are used for the adjustment of later project energy consumption.

[0116] The binary pkl file exported from the building energy consumption prediction model obtained by training the XGBoost model and the dat file storing data will be used for building energy consumption calculation in the multi-objective optimization process.

[0117] 4. Multi-objective optimization design:

[0118] For the coupled-wall modular prefabricated steel structure ultra-low energy consumption building of the present invention, square steel tube section columns are used for columns, and hot-rolled H-shaped steel is selected for beams. In the BIM parametric model, the parameters that need to be optimized include 13 variable parameters such as the total building length, total building width, number of building floors, column size, beam size, exterior wall insulation layer thickness, roof insulation layer thickness, ground insulation layer thickness, east-facing window-wall ratio, west-facing window-wall ratio, south-facing window-wall ratio, north-facing window-wall ratio, and exterior window type. In order to ensure the objectivity of the optimization, the building cost per unit area and the building energy consumption per unit area are used as the optimization objective functions. The calculation models for the building cost per unit area and the building energy consumption per unit area are respectively constructed:

[0119] Cost(x) = TC / (L * W * N)

[0120] Energy(x) = E / (L * W * N)

[0121] Where: TC - Residential building economic cost

[0122] L - Total building length;

[0123] W - Total building width;

[0124] N - Total number of building floors;

[0125] E - Total building energy consumption, which is obtained by calculating based on the building energy consumption prediction model determined by the machine learning algorithm.

[0126] Residential building economic cost:

[0127]

[0128] Where: C b — The cost of the foundation wall;

[0129] C win — The cost of the windows;

[0130] C r — The cost of the roof insulation;

[0131] C w — The cost of the external wall insulation;

[0132] C d — The cost of the ground insulation;

[0133] r — The discount rate;

[0134] y i ,z i — The replacement life;

[0135] C s — The system cost;

[0136] C 钢 — The cost of the steel frame, which is the steel consumption obtained from parametric modeling * the unit price of steel.

[0137] The thermal coefficient limit standards of the peripheral enclosure structures stipulated in the "Energy Efficiency Standard for Residential Buildings in Hebei Province" are used as the constraint conditions for the external wall insulation layer, roof insulation layer, ground insulation layer, window-wall ratio of each orientation, and the type of external windows, and have been considered within the range of parameter changes. The constraint conditions for the steel frame structure are verified by referring to the "Code for Design of Steel Structures" for important indicators such as the strength, stiffness, and stability of flexural-compression members (columns) and flexural members (beams). The steel consumption of the steel frame structure is determined during dynamic layout. After multi-objective optimization, all the optimal solution sets on the Pareto front need to fully meet the following constraint conditions of the steel frame.

[0138]

[0139]

[0140] v b ≤ [v b = I b / 400

[0141] λ cx = l cx / i cx ≤ [λ]; λ cy = l cy / i cy ≤ [λ]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147] 1×10 5 ≤I b ≤5×10 9

[0148] 1×10 6 ≤I c1 ,I c2 ≤5×10 9

[0149] Where: M bx —Design value of bending moment about the x-axis at the same cross-section;

[0150] W bx —Net section modulus about the x-axis;

[0151] γ x —Section plastic development coefficient about the main axis x;

[0152] f—Design value of flexural and compressive strength of steel;

[0153] N c —Design value of axial force of the column (N);

[0154] A c —Cross-sectional area of the column (mm 2 );

[0155] W cx —Section modulus of the column (mm 3 );

[0156] l cx 、l cy —Calculated lengths of the column about the x-axis and y-axis;

[0157] β mx ,β tx —Equivalent bending moment coefficients for in-plane and out-of-plane stability checks;

[0158] —Stability coefficients of axially compressed members in the plane of bending moment and out of the plane of bending moment;

[0159] Δ T —Hypothetical displacement of the structure vertex.

[0160] Based on the multi-objective optimization algorithm NSGA-II, the multi-objective optimization design of the coupled-wall modular prefabricated steel structure ultra-low energy consumption building is carried out, and the specific process is as Figure 5 shown. After the dual-objective problem model of the prefabricated steel structure ultra-low energy consumption building is established and the value range of the optimization parameters is determined, the optimization parameters are encoded to generate the initial population and the evolutionary generation Gen = 1 is set. If the first-generation sub-population is generated, then Gen = 2; otherwise, selection, crossover, and mutation operations are performed on the initial population, and then the parent population and the offspring population are merged to generate a new population. If a new population is generated, the objective function of its individuals is calculated, and operations such as non-dominated sorting and elitist strategy are performed; otherwise, genetic operations are performed on the new population. If Gen ≥ the set value or the convergence condition is reached, the algorithm ends; otherwise, it will loop and calculate from the new population merged by the parent and offspring. Finally, the Pareto optimal solution set of "cost + energy consumption" is output, and the front distribution of the Pareto solution set is as Figure 6 shown. The solution corresponding to point B is the solution with the most obvious optimization of the building cost per unit area, the solution corresponding to point C is the solution with the most obvious optimization of the building energy consumption per unit area, and the solution corresponding to point A is the solution with the best building cost per unit area and building energy consumption per unit area.

[0161] It can be seen from the front distribution diagram of the Pareto solution set that different value combinations of the design parameters of the coupled-wall modular building will produce different index values of the building cost per unit area and the building energy consumption per unit area, and there is a mutually exclusive relationship of one increasing while the other decreasing between the two objective values. The reason is that the better the material performance of the coupled-wall modular building, the smaller the building energy consumption per unit area, and correspondingly, the higher the building cost per unit area.

[0162] According to the obtained Pareto optimal solution set scheme, calculate the distribution frequency of each parameter value in the optimal solution set scheme to study the value law of the parameters in the optimal solution set scheme, as Figure 7 shown. Except for the thickness of the external wall insulation layer and the thickness of the roof insulation layer, the remaining parameters have optimal values within the optimization search range, which is consistent with the conclusion obtained from the feature importance analysis based on the XBGoost building energy consumption prediction model, further enhancing the credibility and accuracy of the artificial intelligence algorithm and the multi-objective optimization algorithm.

[0163] The multi-objective optimization module is encapsulated in the programmable module Python Script in the parametric modeling process. By selecting the scheme in the Pareto solution set, the corresponding BIM model can be directly generated, and finally, the design scheme of the coupled-wall modular prefabricated steel structure ultra-low energy consumption building that meets the actual project cost and energy consumption is completed.

[0164] It should be emphasized that the embodiments described in the present invention are illustrative rather than restrictive. Therefore, the present invention includes, but is not limited to, the embodiments described in the specific implementation manners. Any other implementation manners derived by those skilled in the art based on the technical solutions of the present invention also fall within the scope of protection of the present invention.

[0165] Matters not described in the present invention apply to the prior art.

Claims

1. A multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings, considering two aspects of the cost and energy consumption of prefabricated steel structure ultra-low energy consumption buildings, characterized in that, The specific process of this design method is as follows: 1) Parametric modeling: Since the prefabricated steel structure ultra-low energy consumption building has the modular and standardized characteristics of integrated room module units, different building total lengths and different building total widths have different household layout methods. Fully considering the cantilever minimization, parametric layout principle and the structural assembly method at the construction site, automatically match the household layout of the current building total length and building total width; Generate a prefabricated steel structure building that meets specific goals by adjusting different building total lengths and building total widths. By adjusting the number of building floors, the household layout of the entire prefabricated steel structure can be completed, and then assign material properties to the beams and columns to obtain the BIM model of the steel frame; Conduct structural stability calculations on the BIM model of the steel frame according to the specifications, and count the total steel consumption of the steel structure frame in the BIM model of the steel frame. Obtain the steel frame cost in the total cost calculation according to the steel unit price; 2) Analysis of influencing factors of building energy consumption: Use building energy consumption simulation software to analyze the building energy consumption changes of the performance parameters of the peripheral envelope structure materials that affect building energy consumption one by one, and determine the types and respective value ranges of the key peripheral envelope structure material performance parameters that affect building energy consumption; 3) Building energy consumption prediction: Construct a building energy consumption database based on the types and respective value ranges of the key peripheral envelope structure material performance parameters determined by the building energy consumption influencing factor analysis module. At the same time, use the building energy consumption database to train and evaluate various intelligent algorithms in the artificial intelligence algorithm database, and use the intelligent algorithm with the optimal performance as the building energy consumption prediction model; 4) Multi-objective optimization Use the building total length, building total width, total number of building floors, column size, beam size, and key peripheral envelope structure material performance parameters to form a BIM parametric library. Based on the data in the BIM parametric library, use the unit area building cost Cost(x) and the unit area building energy consumption Energy(x) as the optimization objective functions, and construct a multi-objective optimization design calculation model according to the following formula; Cost(x) = TC / (L * W * N) Energy(x) = E / (L * W * N) Where, TC - the economic cost of residential buildings; L - the total length of the building; W - the total width of the building; N - the total number of building floors; E - the total building energy consumption is obtained from the building energy consumption prediction model in step 3); The economic cost TC of residential buildings: Among them: C b — Basic wall cost; C win — Window cost; C r — Roof insulation cost; C w — Exterior wall insulation cost; C d — Floor insulation cost; r — Discount rate; y i ,z i — Replacement life; C s — System cost; C 钢 — Steel frame cost, obtained by multiplying the steel consumption obtained in the parametric modeling in step 1) by the unit price of steel; The constraint conditions for multi-objective optimization are the thermal coefficient limit standard constraints of the key peripheral envelope structure material performance parameters and the strength, stiffness and stability constraints of the steel frame structure; The strength, stiffness and stability constraints of the steel frame structure refer to meeting the following conditions: v b ≤ [v b = L / 400 λ cx = l cx / i cx ≤ [λ]; λ cy = l cy / i cy ≤ [λ] 1×10 5 ≤I b ≤5×10 9 1×10 6 ≤I c1 ,I c2 ≤5×10 9 Where: M bx — Design value of bending moment about the x-axis at the same cross-section; W bx — Net sectional modulus about the x-axis; γ x — Section plastic development coefficient about the main axis x; f — Design value of steel's flexural and compressive strength; N c — Design value of axial force of the column, unit N; A c — Cross-sectional area of the column, unit mm 2 ; W cx — Section modulus of the column, unit mm 3 ; l cx 、l cy — Calculated lengths of the column about the x-axis and y-axis; β mx , β tx — Equivalent bending moment coefficients for stability checks in-plane and out-of-plane; — Stability coefficients of axially compressed members in-plane and out-of-plane of the bending moment; Δ T — Hypothetical displacement at the top of the structure; x, y are the x-axis and y-axis during the BIM model modeling of the steel frame; Obtain the optimal solution using the multi-objective optimization algorithm, put the optimal solution into the BIM model of the steel frame in the parametric modeling process, and obtain the final BIM model to realize the optimized design of the prefabricated steel structure ultra-low energy consumption building.

2. The multi-objective optimization design method for prefabricated steel structure ultra-low energy consumption buildings according to claim 1, characterized in that, The key peripheral envelope structure material performance parameters include the thickness of the exterior wall insulation layer, the thickness of the roof insulation layer, the thickness of the ground insulation layer, the east window-wall ratio, the west window-wall ratio, the south window-wall ratio, the north window-wall ratio, and the type of exterior window.

3. The multi-objective optimization design method for a prefabricated steel structure ultra-low energy consumption building according to claim 1, characterized in that An energy consumption prediction model for buildings is established using a variety of artificial intelligence algorithms, and the mean absolute error and the coefficient of determination R 2 are used as evaluation indicators for the accuracy performance of the machine learning model for evaluation.

4. An integrated steel structure ultra-low energy consumption building multi-objective optimization design platform, characterized in that The design platform adopts the method described in claim 1, including: A parametric modeling module, which is used to determine the layout mode of the housing type of the prefabricated steel structure according to the characteristics of the prefabricated steel structure ultra-low energy consumption building, and establish a BIM model of the steel frame according to the material properties of the beam and column; A building energy consumption influencing factor analysis module, which is used to determine the types and respective value ranges of the key envelope structure material performance parameters that affect building energy consumption by using building energy consumption simulation software; An artificial intelligence algorithm database, which contains various intelligent algorithms; A building energy consumption prediction module, which constructs a building energy consumption database based on the types and respective value ranges of the key envelope structure material performance parameters determined by the building energy consumption influencing factor analysis module, and at the same time uses the building energy consumption database to train and evaluate various intelligent algorithms in the artificial intelligence algorithm database, and uses the intelligent algorithm with the optimal performance as the building energy consumption prediction model; A multi-objective optimization module, which forms a BIM parameter library with the total length of the building, the total width of the building, the total number of building floors, the column size, the beam size, and the key envelope structure material performance parameters. Based on the data in the BIM parameter library, the unit area building cost Cost(x) and the unit area building energy consumption Energy(x) are used as the objective functions of the optimization to perform multi-objective optimization calculations to determine the optimal solution set; the steel frame cost calculated from the steel consumption in the BIM model of the steel frame is considered in the unit area building cost Cost(x), and the constraint conditions of the multi-objective optimization are the thermal coefficient limit standard constraints of the key envelope structure material performance parameters and the strength, stiffness, and stability constraints of the steel frame structure; A scheme generation module, which applies the scheme in the optimal solution set obtained by the multi-objective optimization module to the BIM model of the steel frame in the parametric modeling module to obtain the final BIM model.

5. The multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings according to claim 4, characterized in that The design platform also includes a structural stability calculation module, which is used to calculate the stability of the BIM model of the steel frame.

6. The multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings according to claim 4, characterized in that The parametric modeling module is implemented through the Dynamo visualization programming module, and the multi-objective optimization module is implemented through the programmable module Python Script in the Dynamo visualization programming.

7. The multi-objective optimization design platform for prefabricated steel structure ultra-low energy consumption buildings according to any one of claims 4-6, characterized in that, The design platform is used for the design of the coupled-column modular prefabricated steel structure ultra-low energy consumption building.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it can implement the steps of the multi-objective optimization design method of the prefabricated steel structure ultra-low energy consumption building described in any one of claims 1-3.

Images