Geometric design optimization method and system

Abstract

The invention belongs to the field of geometric design and optimization, provides a geometric design optimization method and system, and aims to improve the accuracy of data processing by disassembling a geometric model in combination with geometric features. Geometric design parameters are set according to geometric feature types, so that the accuracy, reliability and scientificity of the design parameters are ensured; by selecting whether to reconstruct the geometric model or not, different modes are selected to optimize the geometric model, so that potential optimal geometry can be deeply excavated; geometric consistency detection is carried out on the reconstructed geometric model, so that the integrity and correctness of the geometric model are ensured; the values of the geometric design parameters are optimized through an optimization algorithm, and it can be guaranteed that the assignment of each geometric feature meets a set standard; the number of geometric design parameters is adjusted through an optimization algorithm, and the working efficiency is improved; performance evaluation is carried out on the optimized geometric model, a quantitative index basis is defined, and the reliability and practicability of the geometric model are improved.

Description

Technical Field

[0001] The present invention relates to the field of geometric design and optimization, and particularly to a geometric design optimization method and system. Background Art

[0002] Geometric design and optimization is an important research field in modern engineering design, involving the intersection and integration of multiple disciplines such as structure, fluid, and heat conduction. With the development of computer technology and numerical calculation methods, the research on geometric design optimization has gradually deepened, well replacing the traditional manual optimization method, and its application scope is also increasingly wide.

[0003] Although significant progress has been made in geometric design optimization, many challenges still remain. During the design process, limited by the experience or thinking of engineers, the setting of optimization parameters is not comprehensive enough to cover the actual situation, and the geometric design has poor adaptability, resulting in the boundaries of optimization parameters not conforming to actual physical phenomena. In the process of rapid iteration, it cannot meet the needs of actual engineering. In addition, the complexity of optimization problems often leads to high computational costs, and the geometric model is often simplified during the process, which will cause the loss of characteristic phenomena, thus affecting the accuracy and reliability of the optimized structure. Summary of the Invention

[0004] In order to solve the problems that in the existing geometric design optimization process, the simplification of design parameters leads to the loss of characteristic phenomena, affecting the accuracy and reliability of the optimized structure, and the optimized design parameters do not meet the needs of actual engineering, a geometric design optimization method and system are proposed, and the model is optimized and adjusted to ensure that the optimized design parameters can meet the needs of actual engineering.

[0005] The solution of the present invention is as follows:

[0006] A geometric design optimization method includes the following steps:

[0007] S1: Import the initial geometry and capture geometric features: Import the initial geometric model; capture the geometric features of the initial geometric model, and set geometric design parameters according to the geometric feature types.

[0008] S2: Select an optimization loop and perform optimization: Select an optimization loop based on whether to reconstruct the geometric model according to the geometric design parameters in S1; optimize the geometric design parameters of the initial geometric model in S1 based on the optimization loop to obtain an optimized geometric model.

[0009] S21: Select an optimization loop: Select an optimization loop based on whether to reconstruct the geometric model according to the geometric design parameters in S1; the parameter optimization loop includes: a loop for reconstructing the geometric model and a loop for not reconstructing the geometric model.

[0010] S22: Loop for non-reconstructing geometric model: Based on the optimization algorithm, adjust the numerical values of the geometric design parameters described in S1 to meet the set threshold of the first evaluation index, and obtain the numerical values of the geometric design parameters that meet the optimization objective; perform geometric deformation on the initial geometric model through the obtained numerical values of the geometric design parameters, and output the optimized geometric model.

[0011] S23: Loop for reconstructing geometric model: Perform full parametric modeling using all the geometric design parameters in S1 to obtain a reconstructed geometric model, and perform consistency detection based on the second evaluation index; the second evaluation index is to compare the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model; if the result of the consistency detection is no, adjust the numerical values or quantities of the geometric design parameters in combination with the optimization algorithm, and repeat the operation in step S23; if the result of the consistency detection is yes, output the optimized geometric model.

[0012] S3: Performance evaluation and output of geometric model: Perform performance evaluation on the optimized geometric model output by S2 based on the first evaluation index; if the evaluation fails, jump to step S2; when the adjustment iteration times reach the threshold and the evaluation still fails, output the geometric model marked as unqualified in performance evaluation; if the evaluation passes, output the geometric model marked as qualified in performance evaluation.

[0013] Preferably, the step of importing the initial geometry and capturing geometric features is as follows:

[0014] S11: Decompose the initial geometric model based on geometric features to obtain corresponding geometric features.

[0015] S12: Set geometric design parameters for the geometric features according to the types of the geometric features obtained in S11; the types of the geometric features include: points, lines, and surfaces; for the points, set the geometric design parameters in the form of coordinates; for the lines, first identify the type of the line, and then set the geometric design parameters according to the type of the line; for the surfaces, first identify the type of the surface, and then set the geometric design parameters according to the type of the surface.

[0016] Preferably, the first evaluation index is a mechanical performance index generated based on the force analysis of the geometric model in a specific scenario; the mechanical performance index includes but is not limited to: resistance at different positions, thrust at different positions, lift at different positions, lift-to-drag ratio at different positions.

[0017] Preferably, after adjusting the numerical values of the geometric design parameters through the optimization algorithm, the adjusted numerical values should also meet the set range of the geometric design parameters.

[0018] Preferably, the steps of the loop for reconstructing the geometric model are as follows:

[0019] B1: Perform full parametric modeling based on geometric design parameters to obtain a reconstructed geometric model;

[0020] B2: Perform consistency detection on the reconstructed geometric model based on the second evaluation index;

[0021] B3: If the consistency detection result is no and the number of numerical iteration optimizations has not reached the threshold, adjust the values of the geometric design parameters through an optimization algorithm and repeat step B2; when the number of numerical iteration optimizations reaches the threshold and the consistency detection result is no, adjust the number of geometric design parameters through an optimization algorithm and repeat steps B1 - B2; if the consistency detection result is yes, or when the reconstruction iteration number reaches the threshold and the consistency detection result is no, output the optimized geometric model.

[0022] Preferably, when adjusting the values of the geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the values of the geometric design parameters with the goal of minimizing the tolerance of the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model; when adjusting the number of geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the number of geometric design parameters with the goal of minimizing the number of geometric design parameters.

[0023] Preferably, the step of performing consistency detection on the reconstructed geometric model based on the second evaluation index is as follows:

[0024] C1: Perform meshing on the initial geometric model and the reconstructed geometric model based on three - dimensional profiles and three directions of the three - dimensional coordinate system to obtain the section lines of the initial geometric model and the reconstructed geometric model at the three - dimensional profile positions;

[0025] C2: Compare the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model at the three - dimensional profile positions. When the tolerance of the coincidence degree is within the manually set consistency tolerance range, it is determined that the initial geometric model and the reconstructed geometric model are consistent.

[0026] Preferably, the method of performance evaluation is:

[0027] D1: Set the first evaluation index according to the application field of the initial geometric model and set the compliance threshold of the first evaluation index;

[0028] D2: Calculate the first evaluation index described in D1 of the optimized geometric model through numerical simulation technology and compare the calculation result with the compliance threshold described in D1; if the calculation result reaches the compliance threshold described in D1, the performance evaluation of the optimized geometric model meets the standard.

[0029] A geometric design optimization system, comprising:

[0030] Input module: used to import the geometric model that needs to be geometrically optimized;

[0031] Geometric feature capture module, including: a geometric feature capture unit for capturing geometric features of the geometric model imported from the input model; a design parameter setting unit for setting geometric design parameters of the geometric features captured by the geometric feature capture unit;

[0032] Optimization selection module; used to select whether to implement optimization by means of re - modeling the geometric model; if it is selected to optimize the geometric model by adjusting the values of geometric design parameters, it jumps to the geometric optimization module; if it is selected to optimize the geometric model by re - modeling the geometric model, it jumps to the geometric reconstruction optimization module;

[0033] Geometric optimization module: a parameter optimization unit for adjusting the values of geometric design parameters using an optimization algorithm; a geometric deformation unit for updating the initial geometric model form according to the values of geometric design parameters output by the parameter optimization unit; an output unit for outputting the geometric model updated by the geometric deformation unit;

[0034] Geometric reconstruction optimization module, including: a full - parametric modeling unit for constructing a geometric model based on the geometric design parameters output by the design parameter setting unit or the design parameter quantity adjustment unit; a consistency detection unit for comparing the coincidence degree between the geometric model output by the full - parametric modeling unit and the geometric model imported by the input module; a judgment unit for judging whether the number of numerical iteration optimizations exceeds the set maximum number of iterations if the consistency detection result is negative; a design parameter value optimization unit for adjusting the values of geometric design parameters using an optimization algorithm if the result of the judgment unit is negative; a design parameter quantity adjustment unit for adjusting the quantity of geometric design parameters using an optimization algorithm if the result of the judgment unit is positive; an output unit for outputting the geometric model when the result of the consistency detection unit is positive;

[0035] Performance evaluation module: a performance evaluation unit for performing performance evaluation on the geometric model output by the output unit in the geometric optimization module or the geometric reconstruction optimization module; if the evaluation fails, it jumps to the optimization selection module to perform the next geometric model adjustment and optimization process;

[0036] Geometric library: used to collect the geometric models and the values of geometric design parameters that pass the evaluation by the performance evaluation module.

[0037] Beneficial effects:

[0038] The present invention provides a geometric design optimization method and system. By disassembling the geometric model in combination with geometric features, the accuracy of data processing is improved. By setting geometric design parameters according to the types of geometric features, the accuracy, reliability, and scientificity of the design parameters are ensured, and at the same time, the efficiency of setting design parameters can be significantly improved. By choosing whether to perform geometric model reconstruction and selecting different ways to optimize the geometric model, it is convenient to deeply explore potential optimal geometries. By performing geometric consistency detection on the reconstructed geometric model, it is ensured that the reconstructed geometric model will not have geometric distortion, and the integrity and correctness of the geometric model are ensured, thereby enhancing the reliability of the design. In the loop where the geometric model is not reconstructed, by optimizing the numerical values of the geometric design parameters through an optimization algorithm, the geometric model can more accurately meet the preset performance indicators. In the loop where the geometric model is reconstructed, by optimizing the numerical values of the geometric design parameters through an optimization algorithm, the numerical values of the geometric design parameters of the reconstructed geometric model can be precisely adjusted to improve its consistency with the initial geometric model. At the same time, by optimizing the values of the geometric design parameters through an optimization algorithm, it is ensured that the assignment of each geometric design parameter conforms to the set standard, avoiding the problem of inconsistent assignment results due to differences in understanding in manual assignment. By adjusting the number of geometric design parameters through an optimization algorithm, the work efficiency is improved, and at the same time, mistakes that may be caused by human operations are avoided. By performing performance evaluation on the optimized geometric model, the basis for quantitative indicators is clarified, and the reliability and practicality of the geometric model are enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flowchart of a geometric design optimization method.

[0040] Figure 2 is a framework diagram of a geometric design optimization system. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0041] The present invention will be further described below with reference to the drawings and embodiments.

[0042] Embodiment 1:

[0043] As Figure 1 shown, a geometric design optimization method specifically includes the following steps:

[0044] S1: Import the initial geometry and capture geometric features: Import the initial geometric model; capture the geometric features of the initial geometric model, and set geometric design parameters according to the types of the geometric features;

[0045] S2: Select the optimization loop and perform optimization: Select the optimization loop based on whether to perform geometric model reconstruction according to the geometric design parameters described in S1; optimize the geometric design parameters of the initial geometric model described in S1 based on the optimization loop to obtain an optimized geometric model;

[0046] S21: Select an optimized loop: Select an optimized loop based on whether to reconstruct the geometric model according to the geometric design parameters described in S1; the parameter optimization loop includes: a loop for reconstructing the geometric model and a loop for not reconstructing the geometric model.

[0047] S22: Loop for not reconstructing the geometric model: Based on an optimization algorithm, adjust the values of the geometric design parameters described in S1 to meet the set threshold of the first evaluation index, and obtain the values of the geometric design parameters that meet the optimization goal; perform geometric deformation on the initial geometric model through the obtained values of the geometric design parameters, and output the optimized geometric model.

[0048] S23: Loop for reconstructing the geometric model: Perform full-parametric modeling using all the geometric design parameters in S1 to obtain a reconstructed geometric model, and perform consistency detection based on the second evaluation index; the second evaluation index is to compare the coincidence degree of the sectional lines of the initial geometric model and the reconstructed geometric model; if the result of the consistency detection is no, combine the optimization algorithm to adjust the values or quantity of the geometric design parameters, and repeat the operation in step S23; if the result of the consistency detection is yes, output the optimized geometric model.

[0049] S3: Performance evaluation and output of the geometric model: Perform performance evaluation on the optimized geometric model output by S2 based on the first evaluation index; if the evaluation fails, jump to step S2; when the adjustment iteration times reach the threshold and the evaluation still fails, output the geometric model marked as unqualified in performance evaluation; if the evaluation passes, output the geometric model marked as qualified in performance evaluation.

[0050] Preferably, the step of importing the initial geometry and capturing geometric features is as follows:

[0051] S11: Decompose the initial geometric model based on geometric features to obtain corresponding geometric features.

[0052] S12: Set geometric design parameters for the geometric features according to the types of the geometric features obtained in S11; the types of the geometric features include: points, lines, and surfaces; for the points, set the geometric design parameters in the form of coordinates; for the lines, first identify the type of the line, and then set the geometric design parameters according to the type of the line; for the surfaces, first identify the type of the surface, and then set the geometric design parameters according to the type of the surface.

[0053] Preferably, the first evaluation index is a mechanical performance index generated based on the force analysis of the geometric model in a specific scenario; the mechanical performance index includes but is not limited to: resistance at different positions, thrust at different positions, lift at different positions, lift-to-drag ratio at different positions.

[0054] Preferably, after adjusting the numerical values of the geometric design parameters through the optimization algorithm, the adjusted numerical values should also meet the set range of geometric design parameters.

[0055] Preferably, the steps of the loop for geometric model reconstruction are as follows:

[0056] B1: Perform full parametric modeling based on the geometric design parameters to obtain a reconstructed geometric model;

[0057] B2: Conduct consistency detection on the reconstructed geometric model based on the second evaluation index;

[0058] B3: If the result of the consistency detection is negative and the number of numerical iteration optimizations has not reached the threshold, adjust the numerical values of the geometric design parameters through the optimization algorithm and repeat step B2; when the number of numerical iteration optimizations reaches the threshold and the result of the consistency detection is negative, adjust the number of geometric design parameters through the optimization algorithm and repeat steps B1 - B2; if the result of the consistency detection is positive, or when the number of reconstruction iterations reaches the threshold and the result of the consistency detection is negative, output the optimized geometric model.

[0059] Preferably, when adjusting the numerical values of the geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the numerical values of the geometric design parameters with the aim of minimizing the tolerance of the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model; when adjusting the number of geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the number of geometric design parameters with the aim of minimizing the number of geometric design parameters.

[0060] Preferably, the steps for conducting consistency detection on the reconstructed geometric model based on the second evaluation index are as follows:

[0061] C1: Perform meshing on the initial geometric model and the reconstructed geometric model based on three - dimensional profiles and three directions of the three - dimensional coordinate system to obtain the section lines of the initial geometric model and the reconstructed geometric model at the three - dimensional profile positions;

[0062] C2: Compare the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model at the three - dimensional profile positions. When the tolerance of the coincidence degree is within the manually set consistency tolerance range, it is determined that the initial geometric model and the reconstructed geometric model are consistent.

[0063] Preferably, the method for performance evaluation is:

[0064] D1: Set the first evaluation index according to the application field of the initial geometric model and set the passing threshold of the first evaluation index;

[0065] D2: Calculate the first evaluation index of the optimized geometric model by numerical simulation technology, and compare the calculation result with the passing threshold described in D1; if the calculation result reaches the passing threshold described in D1, the performance evaluation of the optimized geometric model meets the standard.

[0066] Example 2:

[0067] Here, a specific example is used to illustrate the specific implementation process:

[0068] Import the initial geometry; capture the geometric features of the initial geometry to obtain geometric features; extract corresponding geometric design parameters from the geometric features, and the specific steps are as follows:

[0069] First step, in ship speed performance, a ship shape with less resistance needs to be designed. Decompose the imported initial geometry and extract it based on points, lines, and surfaces respectively. Points are the endpoints of three-dimensional geometry or the intersection points of lines. Lines are three-dimensional curves, including straight lines, arcs, curves, etc. Surfaces can be divided into planes, spheres, cylinders, and free-form surfaces, etc.;

[0070] Second step, set design parameters for the extracted points, lines, and surfaces;

[0071] For points, use coordinates to construct design parameters, a total of N1;

[0072] For lines, there are mainly two processing methods: regular lines and fitted lines;

[0073] Regular lines usually refer to lines with regular shapes. For example, linear curves, etc., can be described by specific expressions, such as straight lines, arcs, hyperbolas, parabolas, etc. Construct a total of N2 design parameters;

[0074] Specifically, for a straight line, it can be determined by its two end points. At this time, the construction of the design parameters of the straight line is transformed into the construction of point design parameters; for a parabola, the construction of the design parameters of this line can be constructed through a parabola function, and at the same time, the starting and ending positions of this section of the parabola are limited, where the starting and ending positions can be transformed into the construction of point design parameters.

[0075] Fitted lines mainly refer to curves obtained through various fitting techniques, usually cannot be described by specific expressions, such as B-Spline curves, NURBS curves, etc. Construct a total of N3 design parameters;

[0076] If the curve is composed of multiple sections of surfaces spliced and integrated, it can be split first, and then the corresponding curve type can be selected for processing;

[0077] For surfaces, there are mainly four processing methods: the first category, elementary analytical surfaces, such as planes, spheres, and cylinders; the second category, free-form surfaces;

[0078] The first category - elementary analytic surfaces, which are surfaces that can be constructed through preliminary analysis, such as planes, spheres, and cylinders;

[0079] A plane is usually formed by planar curves, and its boundary is usually composed of regular curves or fitted curves. The design parameter processing method for lines can be combined. If the plane is a regular geometry (such as an equilateral triangle, isosceles triangle, square, rectangle, regular N-sided polygon, circle, sector, etc.), the design parameters can be optimized, and a total of N4 design parameters can be constructed;

[0080] Specifically, since a plane is composed of one or more curves, the design parameters of the lines in the plane can be constructed by borrowing the design parameters for constructing regular lines and fitted lines, and then the design parameters of the plane can be formed;

[0081] A sphere is usually a spherical surface. A complete sphere can be expressed by the center coordinates and radius of the sphere, but an incomplete sphere can combine the characteristics of the curves on its boundary and the processing method for lines to construct N5 design parameters;

[0082] A cylinder is usually a surface formed by stretching or sweeping an A curve along another or multiple B curves (straight lines, curves, or three-dimensional curves). Its design parameters can be decomposed into the design parameters of the A curve and the design parameters of the B curve, and can be disassembled into the processing method for lines, and a total of N6 design parameters can be constructed;

[0083] The second category - free-form surfaces;

[0084] A free-form surface cannot be simply analyzed like an elementary analytic surface, but is a surface composed of a surface that changes freely in a complex manner, such as the shapes of airplanes, cars, and ships. The curves distributed horizontally and vertically are all fitted lines, and the construction of its design parameters is the most complex. Each horizontal or vertical curve is a spliced or fitted curve, and more design parameters need to be constructed to design the free-form surface, with N7 design parameters constructed;

[0085] If a surface is composed of multiple surfaces spliced and integrated, it can be split first, and then the corresponding surface type can be selected for processing;

[0086] Perform full-parameter modeling for the geometric design parameters set above, and then combine the surface coincidence degree detection method in Patent No. 202311086498.3 to determine the consistency between the reconstructed geometric model and the imported initial geometric model. If they are inconsistent and the number of numerical optimization iterations has not reached the maximum number of iterations, the minimum tolerance of the sectional line coincidence degree between the reconstructed geometric model and the imported initial geometric model is taken as the optimization goal. Adjust the numerical values of the design parameters through optimization algorithms (stochastic gradient descent method, genetic algorithm, Sobel, annealing algorithm, T-search, etc.) to obtain the numerical values of the design parameters that meet the optimization goal; within the range of setting the numerical iteration threshold, if the newly built geometry is consistent with the initial geometry, it means that the design parameters that can characterize the imported ship type are screened out; if, after exceeding the set numerical iteration threshold, the reconstructed geometric model and the initial geometric model still do not reach the tolerance range of consistency, then end the numerical iteration optimization process. Take the minimum number of geometric design parameters as the optimization goal, optimize the number of geometric design parameters through an optimization algorithm, perform full-parameter modeling according to the geometric design parameters after the quantity optimization, and repeat the consistency detection and numerical optimization process.

[0087] For the imported initial ship type, after completing the geometric iteration optimization of the ship type, the resistance performance of the ship can be evaluated through numerical simulation technology; if the resistance performance evaluation fails, the geometric model optimization method can be reselected; by reselecting to remodel the geometric model to change the geometric shape of the ship type, the geometric shape of the ship type can be optimized, and then use numerical simulation technology to evaluate the resistance performance of the reconstructed geometric model that passes the consistency detection for the optimized ship type until the designed resistance performance index is met.

[0088] After obtaining the optimal ship type through performance evaluation, the corresponding geometric model and the geometric design parameters of the corresponding geometric features can be collected into the geometric library.

[0089] As Figure 2 shown, a geometric design optimization system includes:

[0090] Input module: used to import the geometric model that needs geometric optimization;

[0091] Geometric feature capture module, including: a geometric feature capture unit used to capture the geometric features of the geometric model imported from the input model; a design parameter setting unit used to set the geometric design parameters of the geometric features captured by the geometric feature capture unit;

[0092] Optimization selection module; used to select whether to implement optimization by means of remodeling the geometric model; if it is selected to optimize the geometric model by adjusting the numerical values of the geometric design parameters, then jump to the geometric optimization module; if it is selected to optimize the geometric model by means of remodeling the geometric model, jump to the geometric reconstruction optimization module;

[0093] Geometric optimization module: a parameter optimization unit that uses an optimization algorithm to adjust the numerical values of geometric design parameters; a geometric deformation unit that updates the shape of the initial geometric model according to the numerical values of the geometric design parameters output by the parameter optimization unit; an output unit for outputting the geometric model updated by the geometric deformation unit;

[0094] Geometric reconstruction optimization module, including: a full-parametric modeling unit for constructing a geometric model from the geometric design parameters output by the design parameter setting unit or the design parameter quantity adjustment unit; a consistency detection unit for comparing the coincidence degree between the geometric model output by the full-parametric modeling unit and the geometric model imported by the input module; a judgment unit for judging whether the number of numerical iteration optimizations exceeds the set maximum number of iterations if the consistency detection result is negative; a design parameter numerical value optimization unit for adjusting the numerical values of geometric design parameters by an optimization algorithm if the result of the judgment unit is negative; a design parameter quantity adjustment unit for adjusting the quantity of geometric design parameters by an optimization algorithm if the result of the judgment unit is positive; an output unit for outputting the geometric model when the result of the consistency detection unit is positive;

[0095] Performance evaluation module: a performance evaluation unit for performing performance evaluation on the geometric model output by the output unit in the geometric optimization module or the geometric reconstruction optimization module; if the evaluation fails, jump to the optimization selection module to perform the next geometric model adjustment and optimization process;

[0096] Geometric library: used to collect the geometric models and the values of geometric design parameters that pass the evaluation by the performance evaluation module.

[0097] It should be noted that the above specific embodiments can enable those skilled in the art to understand the present invention more comprehensively, but do not limit the present invention in any way. Therefore, although this specification has described the present invention in detail with reference to the drawings and embodiments, those skilled in the art should understand that the present invention can still be modified or equivalently replaced. In short, all technical solutions and their improvements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the patent of the present invention.

Claims

1. A geometric design optimization method, characterized in that: S1: Import the initial geometry and capture geometric features: Import the initial geometric model; Capture the geometric features of the initial geometric model, and set geometric design parameters according to the types of the geometric features; S2: Select the optimization loop and perform optimization: Select the optimization loop based on whether to reconstruct the geometric model according to the geometric design parameters described in S1; Optimize the geometric design parameters of the initial geometric model described in S1 based on the optimization loop to obtain an optimized geometric model; S21: Select the optimization loop: Select the optimization loop based on whether to reconstruct the geometric model according to the geometric design parameters described in S1; The parameter optimization loop includes: a loop for reconstructing the geometric model and a loop for not reconstructing the geometric model; S22: The loop for not reconstructing the geometric model: Adjust the values of the geometric design parameters described in S1 based on the optimization algorithm to make them meet the set threshold of the first evaluation index, and obtain the values of the geometric design parameters that meet the optimization goal; Perform geometric deformation on the initial geometric model through the obtained values of the geometric design parameters, and output the optimized geometric model; S23: The loop for reconstructing the geometric model: Perform full-parametric modeling using all the geometric design parameters in S1 to obtain a reconstructed geometric model, and perform consistency detection based on the second evaluation index; The second evaluation index is to compare the coincidence degree of the section lines of the initial geometric model and the reconstructed geometric model; If the consistency detection result is no, adjust the values or quantities of the geometric design parameters in combination with the optimization algorithm, and repeat the operation in step S23; If the consistency detection result is yes, output the optimized geometric model; S3: Performance evaluation and output of the geometric model: Perform performance evaluation on the optimized geometric model output in S2 based on the first evaluation index; If the evaluation fails, jump to step S2; When the adjustment iteration times reach the threshold and the evaluation still fails, output the geometric model marked as unqualified in performance evaluation; If the evaluation passes, output the geometric model marked as qualified in performance evaluation.

2. A geometric design optimization method according to claim 1, characterized in that, The step of importing the initial geometry and capturing geometric features is: S11: Decompose the initial geometric model based on geometric features to obtain corresponding geometric features; S12: Set geometric design parameters for the geometric features according to the types of the geometric features obtained in S11; The types of the geometric features include: points, lines, and surfaces; For the points, set the geometric design parameters in the form of coordinates; For the lines, first identify the type of the line, and then set the geometric design parameters according to the type of the line; For the surfaces, first identify the type of the surface, and then set the geometric design parameters according to the type of the surface.

3. A geometric design optimization method according to claim 1, characterized in that, The first evaluation index is a mechanical performance index generated based on the force analysis of the geometric model in a specific scenario; The mechanical performance index includes but is not limited to: resistance at different positions, thrust at different positions, lift at different positions, lift-drag ratio at different positions.

4. A geometric design optimization method according to claim 1, characterized in that After adjusting the values of the geometric design parameters through the optimization algorithm, the adjusted values also need to meet the set range of the geometric design parameters.

5. A geometric design optimization method according to claim 1, characterized in that, The steps of the loop for reconstructing the geometric model are: B1: Perform full-parametric modeling based on geometric design parameters to obtain a reconstructed geometric model; B2: Perform consistency detection on the reconstructed geometric model based on the second evaluation index; B3: If the consistency detection result is negative and the number of numerical iteration optimizations has not reached the threshold, adjust the numerical values of the geometric design parameters through an optimization algorithm and repeat step B2; when the number of numerical iteration optimizations reaches the threshold and the consistency detection result is negative, adjust the number of geometric design parameters through an optimization algorithm and repeat steps B1 - B2; If the consistency detection result is positive, or when the reconstruction iteration number reaches the threshold and the consistency detection result is negative, output the optimized geometric model.

6. A geometric design optimization method according to claim 5, characterized in that, When adjusting the numerical values of the geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the numerical values of the geometric design parameters with the tolerance of the coincidence degree of the cutting lines of the initial geometric model and the reconstructed geometric model reaching the minimum as the optimization goal; When adjusting the number of geometric design parameters through the optimization algorithm in step B3, the optimization algorithm adjusts the number of geometric design parameters with the number of geometric design parameters reaching the minimum as the optimization goal.

7. A geometric design optimization method according to claim 5, characterized in that, The step of performing consistency detection on the reconstructed geometric model based on the second evaluation index is: C1: Perform meshing on the initial geometric model and the reconstructed geometric model based on a three-dimensional section and three directions of a three-dimensional coordinate system to obtain the cutting lines of the initial geometric model and the reconstructed geometric model at the three-dimensional section positions; C2: Compare the coincidence degree of the cutting lines of the initial geometric model and the reconstructed geometric model at the three-dimensional section positions. When the tolerance of the coincidence degree is within the manually set consistency tolerance range, it is determined that the initial geometric model and the reconstructed geometric model are consistent.

8. A geometric design optimization method according to claim 1, characterized in that The method for performance evaluation is: D1: Set the first evaluation index according to the application field of the initial geometric model and set the compliance threshold of the first evaluation index; D2: Calculate the first evaluation index described in D1 of the optimized geometric model through numerical simulation technology and compare the calculation result with the compliance threshold described in D1; if the calculation result reaches the compliance threshold described in D1, the performance evaluation of the optimized geometric model meets the standard.

9. A geometric design optimization system formed by the method according to any one of claims 1-8, characterized in that, Including: Input module: Used to import the geometric model that needs geometric optimization; Geometric feature capture module, including: a geometric feature capture unit used to capture the geometric features of the geometric model imported from the input model; a design parameter setting unit used to set the geometric design parameters of the geometric features captured by the geometric feature capture unit; Optimization selection module; Used to select whether to implement optimization by means of remolding the geometric model; if it is selected to optimize the geometric model by adjusting the numerical values of the geometric design parameters, jump to the geometric optimization module; if it is selected to optimize the geometric model by remolding the geometric model, jump to the geometric reconstruction optimization module; Geometric optimization module: a parameter optimization unit that uses an optimization algorithm to adjust the numerical values of the geometric design parameters; a geometric deformation unit that updates the shape of the initial geometric model according to the numerical values of the geometric design parameters output by the parameter optimization unit; an output unit used to output the geometric model updated by the geometric deformation unit; Geometric reconstruction optimization module, including: a full-parameterization modeling unit for constructing a geometric model based on the geometric design parameters output by the design parameter setting unit or the design parameter quantity adjustment unit; a consistency detection unit for comparing the coincidence degree between the geometric model output by the full-parameterization modeling unit and the geometric model imported into the input module; a judgment unit for judging whether the number of numerical iteration optimizations exceeds the set maximum number of iterations if the consistency detection result is negative; a design parameter value optimization unit for adjusting the values of the geometric design parameters through an optimization algorithm if the result of the judgment unit is negative; a design parameter quantity adjustment unit for adjusting the quantity of geometric design parameters through an optimization algorithm if the result of the judgment unit is positive; an output unit for outputting the geometric model when the result of the consistency detection unit is positive; Performance evaluation module: a performance evaluation unit for performing performance evaluation on the geometric model output by the output unit in the geometric optimization module or the geometric reconstruction optimization module; if the evaluation fails, jump to the optimization selection module to perform the next geometric model adjustment and optimization process; Geometric library: used to collect the geometric models and the values of geometric design parameters that pass the evaluation by the performance evaluation module.

Images