Airbag template ice shell building weather resistance optimization design method based on proxy model

Abstract

The invention discloses an air bag template ice shell building weather fastness optimization design method based on an agent model, and relates to the technical field of building design. The method solves the problem that the prior art lacks quantitative evaluation indexes of weather resistance of an ice shell building and lacks construction of an agent model with universality, and comprises the following steps: S1, establishing a microclimate response agent model, generating an ice shell case set through orthogonal experimental design, extracting morphological characteristics and spatial characteristics, and obtaining a wind and snow distribution label in combination with CFD simulation; constructing a neural network agent model; and S2, fusing multiple tools to realize performance optimization, extracting features by utilizing Grasshopper parameterized modeling, calling an agent model to predict surface snow wind distribution, obtaining radiation data in combination with Ladybug, calculating initial and later strain energy and weather resistance indexes in combination with a sublimation rate, and finally performing multi-target optimization on morphological parameters through an NSGA-II algorithm, so as to improve the performance of the surface snow wind distribution. The method has application prospects in the field of ice shell building structure optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of building design, and in particular to a weather resistance optimization design method for an airbag template ice shell building based on an agent model. Background Art

[0002] Currently, the evaluation technology for the weatherability of ice-shell buildings is still in its infancy. The core difficulty stems from the complexity of the performance degradation mechanism of composite ice materials under long-term climatic conditions. While previous studies have revealed computational models for the sublimation rate of ice materials and the dynamic changes in their elastic modulus, a systematic quantitative assessment method for weatherability has yet to be established, resulting in a lack of actionable indicators for weatherability optimization design. Meanwhile, proxy model technology has established a relatively standardized application paradigm in architectural design: Based on Latin hypercube sampling, feature-labeled datasets are extracted and trained using black-box models using algorithms such as artificial neural networks (ANNs) and convolutional neural networks (CNNs), replacing traditional simulation processes to improve efficiency. However, these studies have significant limitations. First, existing research focuses primarily on traditional performance objectives such as morphology and structure, completely lacking models for weatherability prediction. This stems from the lack of quantitative indicators for the weatherability of ice-shell buildings. Second, existing models rely heavily on subjectively defined design parameters and task boundaries, resulting in poor generalization. Repeated sampling and training are required for project transitions, offsetting efficiency gains by the cost of secondary modeling and failing to meet the requirements for cross-project universality. The above-mentioned technical breakpoints have resulted in the long-term dual bottlenecks of uncontrollable weather resistance and low multi-objective coordination efficiency in the optimization of ice shell building performance. Summary of the Invention

[0003] To address the problems in the prior art of lacking quantitative evaluation indicators for the weather resistance of ice shell buildings and lacking a universal proxy model construction, the present invention provides a weather resistance optimization design method for airbag formwork ice shell buildings based on a proxy model, comprising:

[0004] S1. Establish a proxy model of spatial morphological characteristics and microclimate responses;

[0005] S11. Construct a case study of ice shell buildings through orthogonal experimental design to extract morphological and spatial characteristics;

[0006] The morphological features include: Z-axis slope, normalized rib line distance, normalized height, maximum principal curvature and minimum principal curvature; the spatial features include: Y-axis slope and normalized wind direction distance;

[0007] S12, performing CFD simulation of wind environment and snow distribution on the ice shell building case set to obtain label information;

[0008] S13, performing model training based on the morphological features, spatial features and label information to construct an artificial neural network agent model, wherein the artificial neural network includes: 1 input layer, 4 hidden layers and 1 output layer;

[0009] S2. Calculate weather resistance index and conduct multi-objective optimization;

[0010] S21. Define the weather resistance index of ice shell buildings;

[0011] S22. constructing a parametric model of the ice shell building using the Grasshopper platform, and extracting the morphological features and the spatial features;

[0012] S23, calling back the proxy model in S1 through the gh_Cpython plug-in, inputting the morphological features and the spatial features, obtaining snow distribution and wind environment data on the ice shell surface; and combining with the Ladybug plug-in to obtain surface radiation distribution data;

[0013] S24. Based on the snow distribution and wind environment data and the surface radiation distribution data on the ice shell surface, update the morphological structure information, load information, and material performance information of the ice shell building and calculate the sublimation rate;

[0014] S25. Based on the morphological and structural information, load information, and material property information of the ice shell building, as well as the sublimation rate, the strain energy at the initial and later stages of service are calculated using the Karamba3D plug-in, and then a weather resistance index of the ice shell building is calculated.

[0015] S26. Optimizing parameters by using the NSGA-II algorithm of the Wallacei plug-in with the initial strain energy and the weather resistance index as targets.

[0016] Further, in S11, in the morphological features, the Z-axis slope is determined by:

[0017]

[0018] Get, where S Z is the Z-axis slope, N is the number of neighborhood cell faces, is the unit normal vector, is the unit vector in the Z-axis direction;

[0019] The normalized rib line distance is obtained by:

[0020]

[0021] Obtain, among which, is the normalized rib line distance, D j,i is the distance from point j to rib line i, D jis all the shortest distance data;

[0022] The normalized height is obtained by:

[0023]

[0024] Obtain, where H z is the z coordinate data of the point, H j is the z coordinate data of all points;

[0025] The maximum principal curvature and minimum principal pass:

[0026]

[0027] Obtained, where C(v) is the principal curvature tensor, |B| is the area of the vertex's 1-ring neighborhood, β(e) is the angle between the normal vectors of the two sides of edge e, |e∩B| is the length of edge e in neighborhood B, is the edge-wise tensor product.

[0028] Further, in S11, in the spatial feature, the Y-axis slope is obtained by:

[0029]

[0030] Get, where S y is the Y-axis slope, is the unit vector in the y-axis direction;

[0031] The normalized wind direction distance is obtained by:

[0032]

[0033] Obtain, among which, is the normalized wind direction distance, D w D is the vertical distance from the point to the wind inflow boundary surface. n is the distance data of all points.

[0034] Furthermore, in S13, the construction of the artificial neural network agent model is specifically as follows: for snow distribution prediction, a 4-factor model is selected, whose features include the Z-axis slope, the normalized height, the distance from the rib line and the minimum principal curvature; for wind environment prediction, a 7-factor model is selected, which includes the Z-axis slope, the Y-axis slope, the normalized height, the distance from the rib line, the distance from the wind direction, the maximum principal curvature and the minimum principal curvature.

[0035] Furthermore, in S21, the weather resistance index of the ice shell building is:

[0036] W=(U E -U S ) / US

[0037] Obtained, where W is the weather resistance index of the ice shell building, U E is the elastic energy in the initial stage, U S It is the elastic performance for a specified period of service;

[0038] Elasticity can be achieved through:

[0039]

[0040] Get, where F i is the node load, μ i is the displacement of the node.

[0041] Further, in S24, the sublimation rate is:

[0042]

[0043] T s =T a +0.0559q G

[0044] Get, where E FRI is the sublimation rate, v is the wind speed, ω is the paper fiber concentration, H is the cumulative sublimation height, T s is the surface temperature of the material, T a is the air temperature, q G Solar radiation intensity.

[0045] Beneficial effects of the present invention:

[0046] First, compared to existing technologies, this method can establish a generalizable proxy model for airbag formwork iceshell buildings. Its input variables are not influenced by the designer's subjective intentions but are directly extracted from the generated building form and spatial location. Therefore, when applying this method to practical airbag formwork iceshell building projects, pre-trained proxy models can be selected based on the building type. This eliminates the need for repeated training and truly improves the efficiency of the design process.

[0047] Secondly, compared to existing technologies, this invention proposes, for the first time, a quantitative evaluation index for the weather resistance of ice-shell buildings. This enables the implementation of multi-objective optimization techniques in architectural design for the weather resistance of ice-shell buildings, a feat never explored in existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Design system flow chart for the present invention;

[0049] Figure 2 Integrate a technology roadmap for this invention;

[0050] Figure 3 A morphology-performance trade-off comparison diagram for multi-objective optimization of ice shells;

[0051] Figure 4 A comparison chart of model training metrics for snow distribution labels;

[0052] Figure 5 This is the model morphology verification effect diagram for snow distribution labels;

[0053] Figure 6 This is a comparison chart of model training indicators for wind environment labels;

[0054] Figure 7 A comparison chart of model training indicators for wind environment labels. DETAILED DESCRIPTION

[0055] The technical solution of the present invention is further described below with reference to the embodiments, but is not limited thereto. Any modification or equivalent replacement of the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention shall be included in the scope of protection of the present invention. The process equipment or devices not specifically noted in the following examples are all conventional equipment or devices in the art. Unless otherwise specified, the raw materials used in the examples of the present invention can be obtained commercially; unless otherwise specified, the technical means used in the examples of the present invention are all conventional means well known to those skilled in the art.

[0056] Example 1, a method for optimizing the weather resistance of an airbag template ice shell building based on an agent model, comprising:

[0057] S1. Establish a proxy model of spatial morphological characteristics and microclimate responses;

[0058] S11. Construct a case study of ice shell buildings through orthogonal experimental design to extract morphological and spatial characteristics;

[0059] The morphological features include: Z-axis slope, normalized rib line distance, normalized height, maximum principal curvature and minimum principal curvature; the spatial features include: Y-axis slope and normalized wind direction distance;

[0060] S12, performing CFD simulation of wind environment and snow distribution on the ice shell building case set to obtain label information;

[0061] S13, performing model training based on the morphological features, spatial features and label information to construct an artificial neural network agent model, wherein the artificial neural network includes: 1 input layer, 4 hidden layers and 1 output layer;

[0062] S2. Calculate weather resistance index and conduct multi-objective optimization;

[0063] S21. Define the weather resistance index of ice shell buildings;

[0064] S22. constructing a parametric model of the ice shell building using the Grasshopper platform, and extracting the morphological features and the spatial features;

[0065] S23, calling back the proxy model in S1 through the gh_Cpython plug-in, inputting the morphological features and the spatial features, obtaining snow distribution and wind environment data on the ice shell surface; and combining with the Ladybug plug-in to obtain surface radiation distribution data;

[0066] S24. Based on the snow distribution and wind environment data and the surface radiation distribution data on the ice shell surface, update the morphological structure information, load information, and material performance information of the ice shell building and calculate the sublimation rate;

[0067] S25. Based on the morphological and structural information, load information, and material property information of the ice shell building, as well as the sublimation rate, the strain energy at the initial and later stages of service are calculated using the Karamba3D plug-in, and then a weather resistance index of the ice shell building is calculated.

[0068] S26. Optimizing parameters by using the NSGA-II algorithm of the Wallacei plug-in with the initial strain energy and the weather resistance index as targets.

[0069] Specifically, the goal of establishing the proxy model in S1 is to get rid of the influence of the designer's subjective intentions on the usability of the model and achieve efficiency and universality. Therefore, this step does not directly output the prediction results of the optimization target through the building design parameters. In terms of features, they can be extracted from the architectural space form. Although the architectural form is generated by the design parameters, its own characteristics are universal. In terms of labels, wind environment and snow distribution are selected. These two are generally simulated by CFD. They are the most time-consuming steps in obtaining performance indicators and are necessary microclimate environmental factors in the weather resistance mechanism. Therefore, the goal of establishing the proxy model in this step is to predict the surface wind environment and snow distribution through the shell space morphology characteristics.

[0070] In S25, the load information is determined by the snow distribution. The snow-covered part is based on a snow depth of 0.3m and a density of 450kg / m 3 The main parameter of material performance change is the elastic modulus, and the elastic modulus in the later period of service is set to 70% of the initial value. The initial material parameter settings are shown in Table 1 below:

[0071] Table 1

[0072]

[0073] pass Figure 1 and Figure 2It can be seen that the present invention optimizes the morphological structure and material properties simultaneously through proxy models and simulation analysis to achieve comprehensive goals such as minimizing strain energy (structural stability) and improving weather resistance (environmental adaptability).

[0074] Figure 3 The results of S26 parameter optimization are presented. By comparing the trade-off between the initial service strain energy and weather resistance index under different design parameters, the optimal balance point between structural performance and weather resistance of the optimized shell shape can be obtained.

[0075] In S11, in the morphological features, the Z-axis slope is:

[0076]

[0077] Get, where S Z is the Z-axis slope, N is the number of neighborhood cell faces, is the unit normal vector, is the unit vector in the Z-axis direction;

[0078] The normalized rib line distance is obtained by:

[0079]

[0080] Obtain, among which, is the normalized rib line distance, D j,i is the distance from point j to rib line i, D j is all the shortest distance data;

[0081] The normalized height is obtained by:

[0082]

[0083] Obtain, where H z is the z coordinate data of the point, H j is the z coordinate data of all points;

[0084] The maximum principal curvature and minimum principal pass:

[0085]

[0086] Obtained, where C(v) is the principal curvature tensor, |B| is the area of the vertex's 1-ring neighborhood, β(e) is the angle between the normal vectors of the two sides of edge e, |e∩B| is the length of edge e in neighborhood B, is the edge-wise tensor product.

[0087] Specifically, the Z-axis slope is the average of the angles between the normal vectors of the cell faces in the point-ring neighborhood and the z-axis, and is normalized for the slope range of 0° (horizontal plane) to 90° (vertical plane); the normalized rib line distance is the Min-Max normalized data of the shortest distance from the vertex to all rib lines; the normalized height is the Min-Max normalized data of the vertex z coordinate;

[0088] In S11, in the spatial feature, the Y-axis slope is:

[0089]

[0090] Get, where S y is the Y-axis slope, is the unit vector in the y-axis direction;

[0091] The normalized wind direction distance is obtained by:

[0092]

[0093] Obtain, among which, is the normalized wind direction distance, D w D is the vertical distance from the point to the wind inflow boundary surface. n is the distance data of all points.

[0094] Specifically, the Y-axis slope is the average of the angles between the normal vectors of the cell surfaces in the point-ring neighborhood and the y-axis, and the slope range from 0° (same-direction surface) to 180° (opposite-direction surface) is normalized. The normalized wind direction distance is the Min-Max normalized distance from the point to the wind inflow boundary surface:

[0095] In S13, the construction of the artificial neural network agent model is specifically as follows: for snow distribution prediction, a 4-factor model is selected, whose features include the Z-axis slope, the normalized height, the distance from the rib line and the minimum principal curvature; for wind environment prediction, a 7-factor model is selected, which includes the Z-axis slope, the Y-axis slope, the normalized height, the distance from the rib line, the distance from the wind direction, the maximum principal curvature and the minimum principal curvature.

[0096] Specifically, the model training was implemented using the Tensorflow 2.8 framework on the Python 3.9 platform. Taking the airbag template ice shell building as the object, the 5F model (Z-axis slope, normalized height, distance from the rib line, maximum principal curvature, minimum principal curvature), 4F1 model (Z-axis slope, distance from the rib line, maximum principal curvature, minimum principal curvature), 4F2 model (Z-axis slope, normalized height, distance from the rib line, minimum principal curvature), and 3F model (Z-axis slope, distance from the rib line, minimum principal curvature) were set for the snow distribution label. The loss and accuracy of the training process were compared as shown in the figure below. Figure 4shown; through Figure 5 The typical combined spindle-shaped shell morphology was verified and the 4F2 model was found to have the best prediction effect.

[0097] Taking the airbag template ice shell building as the object, the 8F model (Z-axis slope, X-axis slope, Y-axis slope, normalized height, distance from the rib line, distance from the wind direction, maximum principal curvature, minimum principal curvature), 7F1 model (Z-axis slope, Y-axis slope, normalized height, distance from the rib line, distance from the wind direction, maximum principal curvature, minimum principal curvature), 7F2 model (Z-axis slope, X-axis slope, Y-axis slope, normalized height, distance from the rib line, distance from the wind direction, maximum principal curvature, minimum principal curvature) were set for the wind environment label. The training process of the 6F model (Z-axis slope, Y-axis slope, normalized height, distance from the rib line, distance from the wind direction, minimum principal curvature), the 5F1 model (Z-axis slope, Y-axis slope, distance from the rib line, distance from the wind direction), the 5F2 model (Z-axis slope, Y-axis slope, distance from the rib line, distance from the wind direction, minimum principal curvature), and the 4F model (Z-axis slope, Y-axis slope, distance from the rib line, distance from the wind direction) were compared. 2 With MAE Figure 6 As shown, through Figure 7 The typical combined spindle-shaped shell morphology was verified and the 7F1 model was found to have the best prediction effect.

[0098] In S21, the weather resistance index of the ice shell building is:

[0099] W=(U E -U S ) / U S

[0100] Obtained, where W is the weather resistance index of the ice shell building, U E is the elastic energy in the initial stage, U S It is the elastic performance for a specified period of service;

[0101] Elasticity can be achieved through:

[0102]

[0103] Get, where F i is the node load, μ i is the displacement of the node.

[0104] Further, in S24, the sublimation rate is:

[0105]

[0106] T s =T a +0.0559q G

[0107] Get, where E FRI is the sublimation rate, v is the wind speed, ω is the paper fiber concentration, H is the cumulative sublimation height, T s is the surface temperature of the material, T a is the air temperature, q G Solar radiation intensity.

Claims

1. A weather resistance optimization design method for airbag template ice shell buildings based on agent model, characterized in that: include: S1. Establish a proxy model of spatial morphological characteristics and microclimate responses; S11. Construct a case study of ice shell buildings through orthogonal experimental design to extract morphological and spatial characteristics; The morphological features include: Z-axis slope, normalized rib line distance, normalized height, maximum principal curvature and minimum principal curvature; the spatial features include: Y-axis slope and normalized wind direction distance; S12, performing CFD simulation of wind environment and snow distribution on the ice shell building case set to obtain label information; S13, performing model training based on the morphological features, spatial features and label information to construct an artificial neural network agent model, The artificial neural network includes: 1 input layer, 4 hidden layers and 1 output layer; S2. Calculate weather resistance index and conduct multi-objective optimization; S21. Define the weather resistance index of ice shell buildings; S22. constructing a parametric model of the ice shell building using the Grasshopper platform, and extracting the morphological features and the spatial features; S23, calling back the proxy model in S1 through the gh_Cpython plug-in, inputting the morphological features and the spatial features, obtaining snow distribution and wind environment data on the ice shell surface; and combining with the Ladybug plug-in to obtain surface radiation distribution data; S24. Based on the snow distribution and wind environment data and the surface radiation distribution data on the ice shell surface, update the morphological structure information, load information, and material performance information of the ice shell building and calculate the sublimation rate; S25. Based on the morphological and structural information, load information, and material property information of the ice shell building, as well as the sublimation rate, the strain energy at the initial and later stages of service are calculated using the Karamba3D plug-in, and then a weather resistance index of the ice shell building is calculated. S26. Optimizing parameters by using the NSGA-II algorithm of the Wallacei plug-in with the initial strain energy and the weather resistance index as targets.

2. The method for optimizing weather resistance of airbag template ice shell buildings based on agent model according to claim 1 is characterized in that: In S11, in the morphological features, the Z-axis slope is: Get, where S z is the Z-axis slope, N is the number of neighborhood cell faces, is the unit normal vector, is the unit vector in the Z-axis direction; The normalized rib line distance is obtained by: Obtain, among which, is the normalized rib line distance, D j,i is the distance from point j to rib line i, D j is all the shortest distance data; The normalized height is obtained by: Obtain, where H z is the z coordinate data of the point, H j is the z coordinate data of all points; The maximum principal curvature and minimum principal pass: Obtained, where C(v) is the principal curvature tensor, |B| is the area of the vertex's 1-ring neighborhood, β(e) is the angle between the normal vectors of the two sides of edge e, |e∩B| is the length of edge e in neighborhood B, is the edge-wise tensor product.

3. The method for optimizing weather resistance of airbag template ice shell buildings based on proxy model according to claim 1, characterized in that: In S11, in the spatial feature, the Y-axis slope is: Get, where S y is the Y-axis slope, is the unit vector in the y-axis direction; The normalized wind direction distance is obtained by: Obtain, among which, is the normalized wind direction distance, D w D is the vertical distance from the point to the wind inflow boundary surface. n is the distance data of all points.

4. The method for optimizing weather resistance of airbag template ice shell buildings based on agent model according to claim 1, characterized in that: In S13, the construction of the artificial neural network agent model is specifically as follows: for snow distribution prediction, a 4-factor model is selected, whose features include the Z-axis slope, the normalized height, the distance from the rib line and the minimum principal curvature; for wind environment prediction, a 7-factor model is selected, which includes the Z-axis slope, the Y-axis slope, the normalized height, the distance from the rib line, the distance from the wind direction, the maximum principal curvature and the minimum principal curvature.

5. The method for optimizing weather resistance of airbag formwork ice shell buildings based on an agent model according to claim 1, characterized in that: In S21, the weather resistance index of the ice shell building is: W=(U E -U S ) / U S Obtained, where W is the weather resistance index of the ice shell building, U E is the elastic energy in the initial stage, U S It is the elastic performance for a specified period of service; Elasticity can be achieved through: Get, where F i is the node load, μ i is the displacement of the node.

6. The method for optimizing weather resistance of airbag formwork ice shell buildings based on an agent model according to claim 1, characterized in that: In S24, the sublimation rate is determined by: T s =T a +0.0559q G Get, where E FRI is the sublimation rate, v is the wind speed, ω is the paper fiber concentration, H is the cumulative sublimation height, T s is the surface temperature of the material, T a is the air temperature, q G Solar radiation intensity.

Images