CFD (Computational Fluid Dynamics) design method for enhancing multi-throttling air-floating guide rail by machine learning

Abstract

The invention discloses a CFD (Computational Fluid Dynamics) design method for enhancing a multi-throttling air-floating guide rail through machine learning, and relates to the technical field of air-floating guide rail optimization design. Selecting design parameters to establish a design parameter library of the multi-throttling air floating guide rail; establishing a three-dimensional model according to the design scheme, and converting the design parameter library into a three-dimensional model library capable of being used for numerical simulation; establishing grid libraries corresponding to different design schemes; constructing a two-dimensional CFD (computational fluid dynamics) simulation database of the multi-throttling air floating guide rail containing performance characteristics; establishing a design parameter-performance characteristic parameter library by adopting a machine learning method; selecting an optimal design scheme based on machine learning; and iteratively optimizing the design parameters until the optimal design parameters are obtained. The flow field of the three-dimensional guide rail model is simulated through fluid mechanics software Fluent, simulation verification is carried out in combination with CFD, the performance of the multi-throttling air floating guide rail can be effectively predicted, the design scheme can be optimized, the experiment cost is reduced, and the design efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of air-floating guide rail optimization design, and in particular to a machine learning enhanced multi-throttling air-floating guide rail CFD design method. Background Art

[0002] Multi-throttle air-floating guides are of great significance in improving system performance in cutting-edge manufacturing equipment and ultra-precision motion equipment. Due to the complex and high uncertainty of environmental conditions, the design of air-floating guides faces great challenges. Multi-throttle air-floating guides have attracted widespread attention in various high-end mechanical systems due to their high load-bearing capacity and high precision.

[0003] However, since the air-floating guide rail operates under complex airflow conditions, different combinations of design parameters such as hole spacing, hole diameter, intake pressure and grooves will affect the performance of the guide rail. Therefore, it is very important to study the design scheme of the air-floating guide rail and its corresponding mechanical properties under specific working conditions. In addition, the experimental optimization cost of the multi-throttle air-floating guide rail is high and the simulation calculation is large.

[0004] In view of this, there is an urgent need for an implementation plan based on machine learning methods to effectively optimize the guide rail design, thereby avoiding expensive experiments and heavy computational requirements. Summary of the invention

[0005] In order to solve the problems of high experimental cost and large amount of simulation calculation in the selection process of multi-throttle air flotation guide design scheme, the present invention provides a machine learning enhanced multi-throttle air flotation guide CFD design method, which simulates the flow field of the three-dimensional guide model through the fluid mechanics software Fluent, and combines CFD for simulation verification, which can effectively predict the performance of multi-throttle air flotation guides, optimize the design scheme, thereby reducing experimental costs and improving design efficiency.

[0006] To achieve the above object, the present invention adopts the following technical solution: a machine learning enhanced multi-throttle air-floating guide CFD design method, comprising the following steps:

[0007] Step 1: Establish a design parameter library for multi-throttle air flotation guides

[0008] Select design parameters including the transverse dimension L of the air flotation surface, the longitudinal dimension W of the air flotation surface, the number of pore rows N_x, the number of pore columns N_y, the transverse spacing Li of the pores, the longitudinal spacing Wi of the pores, the diameter Φ_Cx-y of the pores, the depth D_Cx-y of the pore grooves, the length L_Gx-y of the pore grooves, the width W_Gx-y of the pore grooves, the depth D_Gx-y of the pore grooves and the pore inlet pressure P0, and establish a design parameter library;

[0009] Step 2: Create a 3D model library for multi-throttle air flotation guides

[0010] According to the design scheme of the multi-throttle air flotation guide rail, a three-dimensional model is established by using a three-dimensional modeling software, and the design parameter library is converted into a three-dimensional model library of the multi-throttle air flotation guide rail that can be used for numerical simulation;

[0011] Step 3: Create a grid library corresponding to different multi-throttle air flotation guide design schemes

[0012] Draw the grids of different design schemes through 3D grid drawing software, generate grid libraries corresponding to different design schemes, prepare basic data required for flow field simulation, and simulate and analyze the flow field of the established 3D model under different working conditions through fluid mechanics analysis software;

[0013] Step 4: Establish a 2D simulation database for multi-throttle air flotation guides

[0014] 4.1. Define the physical model

[0015] A three-dimensional compressible fluid model is used, and the finite volume method is used to solve the turbulence. According to the actual physical situation, the wall slip condition is defined, and the turbulence is set to the Realizable k-ε model. The Sutherland formula is used to calculate the viscosity coefficient of the fluid by combining the turbulent and laminar mixed calculations, and the Reynolds number, friction coefficient, flow velocity and pressure distribution are defined.

[0016] 4.2. Establishing Boundary Conditions

[0017] Apply the air inlet pressure P0 to form the initial throttling at the nozzle, and further throttling through the cavity and groove of the air hole to form an air film between the guide rail and the guide sleeve. Set boundary conditions to ensure that the simulation accurately reflects the air film thickness and flow characteristics;

[0018] 4.3 Numerical Solution

[0019] The least squares iteration method is used to solve the linear and nonlinear terms to ensure that the dynamic changes of the air film thickness and flow characteristics can be accurately captured, and the flow velocity, pressure and shear stress of each node are calculated. The influence of the microstructure composed of the pore grooves and pore trenches on the guide rail performance is comprehensively considered to obtain the pressure distribution and flow characteristics of the air film.

[0020] 4.4 Post-processing analysis

[0021] The bearing capacity W of the air film and the stiffness K_h, rotational stiffness K_θ and total tilting moment M_t of the guide rail are calculated as the performance characteristics of the multi-throttle air-floating guide rail.

[0022] 4.5 Parameter optimization and result analysis

[0023] By changing the parameters of the microstructure, the influence on the bearing capacity W of the air film, the stiffness K_h of the guide rail and the rotational stiffness K_θ is studied, and sensitivity analysis is performed on each design parameter to optimize the design variables to meet the design requirements.

[0024] 4.6. Building a 2D CFD simulation database

[0025] Build a 2D CFD simulation database of multi-throttle air-floating guideway including performance characteristics;

[0026] Step 5: Use machine learning methods to establish a design parameter-performance characteristic parameter library

[0027] By using machine learning methods and based on the parameters generated by the simulation analysis results of the fluid mechanics analysis software, the corresponding relationship between the multi-throttle air-floating guide design scheme and its performance characteristics under different working conditions is established as a design parameter-performance characteristic parameter library for machine learning.

[0028] Step 6: Optimize the design scheme through machine learning methods

[0029] Select the best machine learning method, optimize the design scheme based on machine learning, and select the best design scheme for multi-throttle air flotation guide;

[0030] Step 7: Predict the performance characteristics of multi-throttle air flotation guides and optimize the solution

[0031] 7.1. Based on the finalized machine learning method, predict the design parameters with the best load-bearing capacity W, stiffness K_h and rotational stiffness K_θ under given design parameter conditions;

[0032] 7.2. Use 3D modeling software to build a 3D model of the multi-throttle air flotation guide rail;

[0033] 7.3. Use fluid mechanics analysis software to simulate the flow field near the three-dimensional model of the multi-throttle air-floating guide rail, combine CFD to conduct experimental verification, calculate the performance characteristics of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, and obtain the key performance evaluation indicators of the multi-throttle air-floating guide rail;

[0034] 7.4. Iteratively optimize the design parameters. On the basis of comprehensive consideration of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, optimize the design scheme of the multi-throttle air-floating guide rail, and update the design parameter-performance characteristic parameter library until the optimal design parameters of the multi-throttle air-floating guide rail that meet the performance requirements are obtained.

[0035] Furthermore, the step three includes:

[0036] 3.1 Symmetry simplification and meshing

[0037] Considering the structural symmetry, the 1 / 4 structure symmetrical along the x-axis and z-axis is selected for analysis, the calculation range is simplified to obtain the calculation domain, the calculation domain is discretized using triangular meshes, and the mesh density is selected to ensure a balance between calculation accuracy and efficiency;

[0038] 3.2 Node numbering and unit division

[0039] Number the throttling nodes, internal nodes and boundary nodes, divide the grid according to different unit types, and ensure that the parameters of each unit are correct;

[0040] 3.3 Stiffness matrix calculation and iterative solution

[0041] Calculate the stiffness matrix of each unit and integrate it into the overall stiffness matrix. Introduce the nonlinear equation into the node pressure equation group. Solve the nonlinear equation by the least squares iteration method to calculate the pressure distribution of each node.

[0042] 3.4 Post-processing and mesh optimization

[0043] Calculate static characteristics including unit load capacity, total load capacity and stiffness characteristics, and display them visually. Optimize the calculation accuracy and efficiency of the analysis results by adjusting the grid density.

[0044] 3.5. Write script code

[0045] Write script code to automatically generate grid libraries, generate grids for different design schemes, and establish grid libraries corresponding to different multi-throttling air flotation guide rail design schemes.

[0046] Furthermore, the pore air inlet pressure P0 applied in step 4 does not exceed 0.4 MPa.

[0047] Furthermore, the step six includes:

[0048] 6.1. Select machine learning algorithms and choose the optimal algorithm combination based on the complexity of the problem and the size of the data;

[0049] 6.2. Divide the design parameter-performance characteristic parameter library into training sets and validation sets in proportion, perform data preprocessing, model selection and cross-validation, and perform model training and optimization to ensure the accuracy and robustness of the prediction results of the machine learning model;

[0050] 6.3. Use the commonly used evaluation indicators in machine learning regression algorithms, including root mean square error and mean absolute error, to evaluate the performance of the machine learning model. After the evaluation, select the best machine learning method, optimize the design scheme based on machine learning, and select the best design scheme for the multi-throttle air flotation guide.

[0051] Furthermore, the selection of the machine learning algorithm includes random forest, Gaussian support vector machine, linear support vector machine, gradient boosting and deep learning model.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows: starting from the relationship between different design parameters and the performance characteristics of the air flotation guide rail, the present invention uses machine learning technology to model, simulate and analyze the relationship between different design schemes of the multi-throttle air flotation guide rail and its performance characteristics, and uses the fluid mechanics software Fluent to simulate the flow field of the three-dimensional guide rail model, and combines CFD for simulation verification. It can effectively predict the performance of the multi-throttle air flotation guide rail under different working conditions, can self-iterate and optimize, significantly reduce the computing load, reduce experimental requirements, and comprehensively consider the requirements of key performance indicators such as load-bearing capacity and stiffness. While ensuring the effectiveness of the design scheme, it improves the overall optimization efficiency, providing the possibility for further improving the design performance of the air flotation guide rail. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of the method of the present invention;

[0054] Figure 2 Schematic diagram of a multi-throttle air flotation guide rail when the number of air hole rows N_x=1 and the number of air hole columns N_y=7;

[0055] Figure 3 Schematic diagram of a multi-throttle air flotation guide rail when the number of air hole rows N_x=2 and the number of air hole columns N_y=7;

[0056] Figure 4 Schematic diagram of a multi-throttle air flotation guide rail when the number of air hole rows N_x=2 and the number of air hole columns N_y=6;

[0057] Figure 5 It is a schematic diagram of the structure of the air hole slot and the air hole groove of the multi-throttle air flotation guide;

[0058] Figure 6 It is a parameter diagram of the air hole slot and the air hole groove of the multi-throttle air flotation guide;

[0059] Figure 7 A grid diagram of a design solution. DETAILED DESCRIPTION

[0060] The technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0061] like Figure 1 to Figure 7As shown in the figure, a machine learning enhanced multi-throttle air-floating guide CFD design method is used. Figure 1 As shown, the following steps are included:

[0062] Step 1: Establish a design parameter library for multi-throttle air flotation guides

[0063] Based on the geometric and physical parameters of the multi-throttle air flotation guide, a design parameter library is established by selecting different design parameters. The design parameters of the multi-throttle air flotation guide include the lateral dimension L of the air flotation surface, the longitudinal dimension W of the air flotation surface, the number of pore rows N_x, the number of pore columns N_y, the lateral spacing Li of the pores, the longitudinal spacing Wi of the pores, the diameter Φ_Cx-y of the pore groove, the depth D_Cx-y of the pore groove, the length L_Gx-y of the pore groove, the width W_Gx-y of the pore groove, the depth D_Gx-y of the pore groove and the pore inlet pressure P0. In each design parameter, x represents the corresponding row, y represents the corresponding column, and i represents the number of the two pore spacings symmetrical to the lateral / longitudinal center line of the air flotation surface, which is used for subsequent three-dimensional modeling, mesh drawing and CFD performance analysis.

[0064] In order to understand the design parameters clearly, Figure 2 to Figure 4 As shown, schematic diagrams are shown in turn when the number of pore rows N_x=1, the number of pore columns N_y=7, the number of pore rows N_x=2, the number of pore columns N_y=7, and the number of pore rows N_x=2, the number of pore columns N_y=6, and the pores are numbered according to the number of rows and columns using xy. Figure 2 ( Figure 3 ) is an odd number, and the middle row of pores 1-4 (1-4 and 2-4) is on the longitudinal center line of the air flotation surface, so the middle row of pores is ignored and the transverse spacing of the pores is numbered on both sides. The transverse spacing between pores 1-3 and 1-5 (1-3 and 1-5 and 2-3 and 2-5) is counted as L1, the transverse spacing between pores 1-2 and 1-6 (1-2 and 1-6 and 2-2 and 2-6) is counted as L2, and the transverse spacing between pores 1-1 and 1-7 (1-1 and 1-7 and 2-1 and 2-7) is counted as L3. Figure 2 The number of pore rows N_x = 1 does not exist in the longitudinal spacing of pores. Figure 3 When the number of pore rows N_x=2, the longitudinal spacing between pores is W1. Figure 4 Since the number of pore columns N_y is an even number, the horizontal spacing between pores 1-3 and 1-4 and 2-3 and 2-4 is calculated as L1, the horizontal spacing between pores 1-2 and 1-5 and 2-2 and 2-5 is calculated as L2, the horizontal spacing between pores 1-1 and 1-6 and 2-1 and 2-6 is calculated as L3, and the number of pore rows N_x = 2, then the vertical spacing between pores is calculated as W1. Since the arrangement of pores in each row is consistent and combined with the central symmetry, the horizontal and vertical spacing between each two adjacent pores can be obtained.

[0065] Combination Figure 5-6As shown in the figure, the structure and parameter diagrams of the stomatal groove and stomatal furrow are shown respectively. Figure 5 The pore grooves, pore grooves and pore inlet pressure of the pores are specifically described in the paper. Figure 6 In the figure, the pores in row 1 and column 1 are taken as an example, and their pore groove diameter Φ_C1-1, pore groove depth D_C1-1, pore groove length L_G1-1, pore groove width W_G1-1, and pore groove depth D_G1-1 are respectively represented. Among them, the pore groove length, pore groove width, and pore groove depth can be 0, which means that there is no pore groove in the multi-throttle air flotation guide.

[0066] Step 2: Create a 3D model library for multi-throttle air flotation guides

[0067] According to the design scheme of the multi-throttle air flotation guide rail, a three-dimensional model is established through three-dimensional modeling software such as SolidWorks / ProE, and the design parameter library is converted into a three-dimensional model library of the multi-throttle air flotation guide rail that can be used for numerical simulation.

[0068] Step 3: Create a grid library corresponding to different multi-throttle air flotation guide design schemes

[0069] The meshes of different design schemes are drawn by 3D mesh drawing software ICEM, combined with Figure 7 As shown in the figure, a grid schematic diagram of a design scheme is shown, and finally a grid library corresponding to different design schemes is generated, and the basic data required for flow field simulation is prepared. The flow field of the established three-dimensional model under different working conditions is simulated and analyzed by the fluid mechanics analysis software Fluent, including:

[0070] 3.1 Symmetry simplification and meshing

[0071] Considering the structural symmetry, the 1 / 4 structure symmetrical along the x-axis and z-axis is selected for analysis, and the calculation range is simplified to obtain the calculation domain. Then, the calculation domain is discretized using triangular meshes, and the appropriate mesh density is selected to ensure a balance between calculation accuracy and efficiency.

[0072] 3.2 Node numbering and unit division

[0073] The throttling nodes, internal nodes and boundary nodes are numbered from left to right and from bottom to top. Then, the grid is divided according to different unit types to ensure that the parameters of each unit are correct, which provides a basis for the subsequent calculation of the stiffness matrix.

[0074] 3.3 Stiffness matrix calculation and iterative solution

[0075] Calculate the stiffness matrix of each unit and integrate it into the overall stiffness matrix. Then, introduce the nonlinear equation into the node pressure equation group, solve the nonlinear equation by the least squares iteration method, and calculate the pressure distribution of each node.

[0076] 3.4 Post-processing and mesh optimization

[0077] In the post-processing process, the static characteristics including unit load capacity, total load capacity and stiffness characteristics are calculated and visualized. Then, the calculation accuracy and efficiency of the analysis results are optimized by adjusting the mesh density.

[0078] 3.5. Write script code

[0079] Write script code to automatically generate a grid library, generate grids for different design schemes, improve the degree of automation, and ultimately establish a grid library corresponding to different multi-throttling air flotation guide rail design schemes.

[0080] Step 4: Establish a 2D simulation database for multi-throttle air flotation guides

[0081] 4.1. Define the physical model

[0082] A three-dimensional compressible fluid model is used, and the finite volume method FVM is used to solve the turbulence. According to the actual physical situation, the wall slip condition is defined, and the turbulence is set to the Realizable k-ε model. The Sutherland formula is used to calculate the viscosity coefficient of the fluid by combining the turbulent and laminar mixed calculations, and the Reynolds number, friction coefficient, flow velocity and pressure distribution are defined;

[0083] 4.2. Establishing Boundary Conditions

[0084] Apply the air inlet pressure P0 (not exceeding 0.4MPa) to form the initial throttling at the nozzle, and further throttling through the cavity and groove of the air hole, and finally form an air film between the guide rail and the guide sleeve. Set the boundary conditions to ensure that the simulation accurately reflects the thickness of the air film and the flow characteristics;

[0085] 4.3 Numerical Solution

[0086] The least squares iteration method is used to solve the linear and nonlinear terms to ensure that the dynamic changes of the air film thickness and flow characteristics can be accurately captured, and the flow velocity, pressure and shear stress of each node are calculated. The influence of the microstructure composed of the pore grooves and pore trenches on the guide rail performance is comprehensively considered to obtain the pressure distribution and flow characteristics of the air film.

[0087] 4.4 Post-processing analysis

[0088] The bearing capacity W of the air film is calculated by integrating the pressure distribution, the stiffness K_h of the guide rail is obtained by calculating the differential of the bearing capacity in the direction of the air film thickness, the rotational stiffness K_θ is obtained by calculating the differential of the tilting moment in the direction of the rotation angle θ, and the total tilting moment M_t is calculated according to the moment of each unit on the grid as the performance characteristic of the multi-throttle air-floating guide rail;

[0089] 4.5 Parameter optimization and result analysis

[0090] By changing the parameters of the microstructure, the influence on the bearing capacity W of the air film, the stiffness K_h of the guide rail and the rotational stiffness K_θ is studied, and sensitivity analysis is performed on each design parameter to optimize the design variables to meet the design requirements.

[0091] 4.6. Building a 2D CFD simulation database

[0092] A two-dimensional CFD simulation database of a multi-throttle air-floating guide rail is constructed, including performance characteristics such as load capacity W, stiffness K_h, rotational stiffness K_θ and total tilting moment M_t.

[0093] Step 5: Use machine learning methods to establish a design parameter-performance characteristic parameter library

[0094] Using the machine learning method, according to the parameters generated by the simulation analysis results of the fluid mechanics analysis software Fluent, the corresponding relationship between the multi-throttling air flotation guide design scheme and its performance characteristics under different working conditions is established as the design parameter-performance characteristic parameter library for machine learning. The input data include the design parameters including the lateral dimension L of the air flotation surface, the longitudinal dimension W of the air flotation surface, the number of pore rows N_x, the number of pore columns N_y, the lateral spacing Li of the pores, the longitudinal spacing Wi of the pores, the diameter Φ_Cx-y of the pore groove, the depth D_Cx-y of the pore groove, the length L_Gx-y of the pore groove, the width W_Gx-y of the pore groove, the depth D_Gx-y of the pore groove and the air inlet pressure P0 of the pore. The output data include the performance characteristics including the bearing capacity W, stiffness K_h, rotational stiffness K_θ and the total tilting moment M_t.

[0095] Step 6: Optimize the design scheme through machine learning methods

[0096] 6.1. Select appropriate machine learning algorithms, including but not limited to Random Forest, Gaussian Support Vector Machine (SVM-Gaussian), Linear Support Vector Machine (SVM-Linear), Gradient Boosting and deep learning models (such as 2-layer, 3-layer, 4-layer artificial neural networks ANN20-20, ANN30-30, ANN40-40, etc.), and select the optimal algorithm combination based on the complexity of the problem and the size of the data;

[0097] 6.2. To avoid overfitting during the training of the machine learning algorithm, the design parameter-performance characteristic parameter library is divided into a training set and a validation set in proportion. The recommended ratio is 7:3 or other reasonable ratios. At the same time, necessary data preprocessing is performed, including feature scaling and missing value processing, followed by model selection and cross-validation, and model training and optimization to ensure the accuracy and robustness of the prediction results of the machine learning model.

[0098] 6.3. Use the commonly used evaluation indicators in machine learning regression algorithms, including root mean square error (RMSE) and mean absolute error (MAE) to evaluate the performance of machine learning models. The root mean square error is used to measure the difference between the model's predicted value and the actual value. The result is kept consistent with the original data unit, which is convenient for understanding the model's prediction error; the mean absolute error represents the average of the absolute errors between the estimated value and the actual value, which can directly reflect the prediction accuracy of the model. Generally speaking, the smaller the values ​​of RMSE and MAE, the better the fitting effect of the machine learning model, indicating that the error between the predicted value and the actual value is smaller. The specific formula for the evaluation indicators is as follows:

[0099]

[0100] After evaluation, the best machine learning method is selected, the design scheme is optimized based on machine learning, and the best design scheme for the multi-throttling air flotation guide is selected.

[0101] Step 7: Predict the performance characteristics of multi-throttle air flotation guides and optimize the solution

[0102] 7.1. Based on the finalized machine learning method, predict the design parameters with the best load-bearing capacity W, stiffness K_h and rotational stiffness K_θ under given design parameter conditions;

[0103] 7.2. Use 3D modeling software such as SolidWorks / ProE to build a 3D model of the multi-throttle air flotation guide rail to facilitate subsequent performance analysis;

[0104] 7.3. Use the fluid mechanics analysis software Fluent to simulate the flow field near the three-dimensional model of the multi-throttle air-floating guide rail, combine it with CFD to conduct experimental verification, calculate the performance characteristics of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, and obtain the key performance evaluation indicators of the multi-throttle air-floating guide rail;

[0105] 7.4. Iteratively optimize the design parameters. On the basis of comprehensive consideration of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, optimize the design scheme of the multi-throttle air-floating guide rail, and update the design parameter-performance characteristic parameter library until the optimal design parameters of the multi-throttle air-floating guide rail that meet the performance requirements are obtained.

[0106] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other forms of assembly without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations within the meaning and range of equivalents of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0107] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A machine learning enhanced multi-throttle air-floating guide CFD design method, characterized by: The following steps are involved: Step 1: Establish a design parameter library for multi-throttle air flotation guides Select design parameters including the transverse dimension L of the air flotation surface, the longitudinal dimension W of the air flotation surface, the number of pore rows N_x, the number of pore columns N_y, the transverse spacing Li of the pores, the longitudinal spacing Wi of the pores, the diameter Φ_Cx-y of the pores, the depth D_Cx-y of the pore grooves, the length L_Gx-y of the pore grooves, the width W_Gx-y of the pore grooves, the depth D_Gx-y of the pore grooves and the pore inlet pressure P0, and establish a design parameter library; Step 2: Create a 3D model library for multi-throttle air flotation guides According to the design scheme of the multi-throttle air flotation guide rail, a three-dimensional model is established by using a three-dimensional modeling software, and the design parameter library is converted into a three-dimensional model library of the multi-throttle air flotation guide rail that can be used for numerical simulation; Step 3: Create a grid library corresponding to different multi-throttle air flotation guide design schemes Draw the grids of different design schemes through 3D grid drawing software, generate grid libraries corresponding to different design schemes, prepare basic data required for flow field simulation, and simulate and analyze the flow field of the established 3D model under different working conditions through fluid mechanics analysis software; Step 4: Establish a 2D simulation database for multi-throttle air flotation guides 4.

1. Define the physical model A three-dimensional compressible fluid model is used, and the finite volume method is used to solve the turbulence. According to the actual physical situation, the wall slip condition is defined, and the turbulence is set to the Realizable k-ε model. The Sutherland formula is used to calculate the viscosity coefficient of the fluid by combining the turbulent and laminar mixed calculations, and the Reynolds number, friction coefficient, flow velocity and pressure distribution are defined. 4.

2. Establishing Boundary Conditions Apply the air inlet pressure P0 to form the initial throttling at the nozzle, and further throttling through the cavity and groove of the air hole to form an air film between the guide rail and the guide sleeve. Set boundary conditions to ensure that the simulation accurately reflects the air film thickness and flow characteristics; 4.3 Numerical Solution The least squares iteration method is used to solve the linear and nonlinear terms to ensure that the dynamic changes of the air film thickness and flow characteristics can be accurately captured, and the flow velocity, pressure and shear stress of each node are calculated. The influence of the microstructure composed of the pore grooves and pore trenches on the guide rail performance is comprehensively considered to obtain the pressure distribution and flow characteristics of the air film. 4.4 Post-processing analysis The bearing capacity W of the air film and the stiffness K_h, rotational stiffness K_θ and total tilting moment M_t of the guide rail are calculated as the performance characteristics of the multi-throttle air-floating guide rail. 4.5 Parameter optimization and result analysis By changing the parameters of the microstructure, the influence on the bearing capacity W of the air film, the stiffness K_h of the guide rail and the rotational stiffness K_θ is studied, and sensitivity analysis is performed on each design parameter to optimize the design variables to meet the design requirements. 4.

6. Building a 2D CFD simulation database Build a 2D CFD simulation database of multi-throttle air-floating guideway including performance characteristics; Step 5: Use machine learning methods to establish a design parameter-performance characteristic parameter library By using machine learning methods and based on the parameters generated by the simulation analysis results of the fluid mechanics analysis software, the corresponding relationship between the multi-throttle air-floating guide design scheme and its performance characteristics under different working conditions is established as a design parameter-performance characteristic parameter library for machine learning. Step 6: Optimize the design scheme through machine learning methods Select the best machine learning method, optimize the design scheme based on machine learning, and select the best design scheme for multi-throttle air flotation guide; Step 7: Predict the performance characteristics of multi-throttle air flotation guides and optimize the solution 7.

1. Based on the finalized machine learning method, predict the design parameters with the best load-bearing capacity W, stiffness K_h and rotational stiffness K_θ under given design parameter conditions; 7.

2. Use 3D modeling software to build a 3D model of the multi-throttle air flotation guide rail; 7.

3. Use fluid mechanics analysis software to simulate the flow field near the three-dimensional model of the multi-throttle air-floating guide rail, combine CFD to conduct experimental verification, calculate the performance characteristics of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, and obtain the key performance evaluation indicators of the multi-throttle air-floating guide rail; 7.

4. Iteratively optimize the design parameters. On the basis of comprehensive consideration of the load-bearing capacity W, stiffness K_h and rotational stiffness K_θ, optimize the design scheme of the multi-throttle air-floating guide rail, and update the design parameter-performance characteristic parameter library until the optimal design parameters of the multi-throttle air-floating guide rail that meet the performance requirements are obtained.

2. The machine learning enhanced multi-throttle air-floating guide CFD design method according to claim 1, characterized in that: The step three comprises: 3.1 Symmetry simplification and meshing Considering the structural symmetry, the 1 / 4 structure symmetrical along the x-axis and z-axis is selected for analysis, the calculation range is simplified to obtain the calculation domain, the calculation domain is discretized using triangular meshes, and the mesh density is selected to ensure a balance between calculation accuracy and efficiency; 3.2 Node numbering and unit division Number the throttling nodes, internal nodes and boundary nodes, divide the grid according to different unit types, and ensure that the parameters of each unit are correct; 3.3 Stiffness matrix calculation and iterative solution Calculate the stiffness matrix of each unit and integrate it into the overall stiffness matrix. Introduce the nonlinear equation into the node pressure equation group. Solve the nonlinear equation by the least squares iteration method to calculate the pressure distribution of each node. 3.4 Post-processing and mesh optimization Calculate static characteristics including unit load capacity, total load capacity and stiffness characteristics, and display them visually. Optimize the calculation accuracy and efficiency of the analysis results by adjusting the grid density. 3.

5. Write script code Write script code to automatically generate grid libraries, generate grids for different design schemes, and establish grid libraries corresponding to different multi-throttling air flotation guide rail design schemes.

3. The machine learning enhanced multi-throttle air-floating guide CFD design method according to claim 1, characterized in that: The pore air inlet pressure P0 applied in the step 4 does not exceed 0.4 MPa.

4. The machine learning enhanced multi-throttle air-floating guide CFD design method according to claim 1, characterized in that: The step six comprises: 6.

1. Select machine learning algorithms and choose the optimal algorithm combination based on the complexity of the problem and the size of the data; 6.

2. Divide the design parameter-performance characteristic parameter library into training sets and validation sets in proportion, perform data preprocessing, model selection and cross-validation, and perform model training and optimization to ensure the accuracy and robustness of the prediction results of the machine learning model; 6.

3. Use the commonly used evaluation indicators in machine learning regression algorithms, including root mean square error and mean absolute error, to evaluate the performance of the machine learning model. After the evaluation, select the best machine learning method, optimize the design scheme based on machine learning, and select the best design scheme for the multi-throttle air flotation guide.

5. The machine learning enhanced multi-throttle air-floating guide CFD design method according to claim 4 is characterized in that: The selection of machine learning algorithms includes random forest, Gaussian support vector machine, linear support vector machine, gradient boosting and deep learning models.

Images