Precise mold design method and system based on intelligent optimization algorithm

Abstract

The invention discloses a precise mold design method and system based on an intelligent optimization algorithm, which can effectively balance a plurality of key performance indexes such as path length, cutting load distribution and surface quality by constructing a multi-objective optimization framework, combining iterative computation to generate a candidate path set and introducing a Pareto optimal solution screening mechanism. On this basis, path parameters are further dynamically adjusted in combination with a process experience database, and path inflection points and transition sections are optimized through a neighborhood disturbance strategy, so that the finally generated processing path not only meets the requirements of high efficiency and stability, but also can adapt to actual processing requirements under different process conditions; according to the method, the machining efficiency and the surface consistency of complex mold parts are remarkably improved, the trial and error cost is reduced, and intelligent and self-adaptive precision mold design and machining path planning are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of mold design, and in particular relates to a precision mold design method and system based on an intelligent optimization algorithm. Background Art

[0002] In the field of precision mold machining, traditional mold design and machining path planning rely primarily on manual experience combined with fixed rules. A geometry-driven approach is typically used to directly generate tool paths based on the part's 3D model. Common methods include contour cutting, offset circular cutting, and reciprocating tool paths. While these methods are simple to implement and widely used for parts with relatively regular structures, they present challenges when faced with complex curved surfaces or high-precision mold parts, such as lengthy paths, uneven cutting forces, and unstable surface quality.

[0003] Some existing solutions incorporate optimization algorithms to improve machining paths, but these often focus on a single objective, such as optimizing path length or reducing cutting force fluctuations, without comprehensive consideration of multiple performance indicators. Furthermore, traditional path optimization methods struggle to dynamically adapt to changes in materials, process parameters, and equipment conditions, making it difficult to balance machining efficiency and quality. Summary of the Invention

[0004] The purpose of the present invention is to provide a precision mold design method and system based on an intelligent optimization algorithm, which improves the processing efficiency and surface consistency of complex mold parts, reduces the trial and error cost, and realizes intelligent and adaptive precision mold design and processing path planning to solve the problem of how to take into account path length, cutting stability and surface quality in the processing of complex mold parts, and realize adaptive adjustment to changes in process conditions, thereby improving the overall processing efficiency and accuracy.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a precision mold design method based on an intelligent optimization algorithm, comprising the following steps:

[0006] The three-dimensional geometric data and processing constraints of the mold parts are collected to generate an initial data set; feature extraction is performed on the initial data set to identify the geometric features and constraint boundaries in the processing path; a multi-objective optimization framework is constructed based on the geometric features and constraint boundaries, and within the multi-objective optimization framework, a set of candidate processing paths is generated through iterative calculation; a multi-dimensional comparative analysis is performed on the candidate processing path set to screen out a Pareto optimal solution set, and based on the Pareto optimal solution set, the path parameters are dynamically adjusted in combination with a process experience database to generate an improved path scheme; a local search is performed on the improved path scheme, and the smoothness of the path inflection points and transition sections is optimized through a neighborhood perturbation strategy to verify the feasibility of the improved path scheme, output the final processing path, and generate CNC code.

[0007] Preferably, the steps include:

[0008] Obtain the three-dimensional geometric model of the mold part and extract the curvature distribution information of each surface;

[0009] Discretize the curvature distribution information into multiple curvature intervals, and count the point density in each interval;

[0010] Combining the point density with the preset sampling step length, calculating the distribution ratio of the number of sampling points in each area;

[0011] According to the sampling point quantity distribution ratio, adaptive sampling is performed in each curvature interval to generate a spatial point cloud set.

[0012] Preferably, the identifying geometric features and constraint boundaries in the machining path includes:

[0013] Performing neighborhood topology analysis on the point cloud set in the initial data set to determine the local surface change trend of each point;

[0014] According to the local surface change trend, extracting the curvature mutation area as the potential geometric feature boundary;

[0015] Combining the geometric feature boundary with the machining direction projection relationship to calculate the infeasible area affected by tool interference;

[0016] The infeasible area and the geometric feature boundary are combined to define the actual available machining path constraint range.

[0017] Preferably, a multi-objective optimization framework is constructed based on the geometric features and constraint boundaries, including:

[0018] Dividing the processing area into multiple sub-areas based on the geometric feature boundaries and constraint range, and marking the surface inclination angle and contact risk level of each sub-area;

[0019] A path planning priority is set for each sub-area. Based on the path planning priority, three parallel optimization objectives are set: minimizing the idle travel distance, balancing the cutting load distribution, and reducing the number of inflection points.

[0020] The priority of each sub-region is mapped to the corresponding target weight to form a target-oriented optimization spatial structure.

[0021] Preferably, generating a set of candidate processing paths through iterative calculation includes:

[0022] Randomly generating a set of initial path sequences within the optimized spatial structure as a first generation path population;

[0023] Performing fitness evaluation on each path in the initial path population, wherein the scoring is based on a weighted sum of path length, load fluctuation amplitude, and number of inflection points;

[0024] Based on the fitness score results, high-scoring paths are selected and a new batch of paths are generated through cross-joining and mutation operations. The position sequence of the new path is generated by a weighted combination of the position sequence of the parent path and the weight coefficient;

[0025] The fitness evaluation and path updating process is repeated until the preset number of iterations is reached, and finally a set of candidate processing paths is output.

[0026] Preferably, screening out the Pareto optimal solution set includes:

[0027] Performing multi-objective performance ranking on each path in the candidate machining path set, and recording the independent ranking of each path under three indicators: path length, cutting force fluctuation, and surface quality;

[0028] Calculate the comprehensive advantage index of each path;

[0029] Sort the paths from high to low according to the comprehensive advantage index, and set a threshold to select the top N paths as the preliminary preferred path group;

[0030] Pairwise comparisons are performed within the preliminary preferred path group, completely dominated paths are excluded, and non-dominated paths are retained to form a Pareto optimal solution set.

[0031] Preferably, the generating of the improved path plan includes:

[0032] Matching each path in the Pareto optimal solution set with historical processing parameters in a process experience database to find historical cases with a similarity higher than a set threshold;

[0033] Based on the tool compensation value and feed rate settings in the historical case, parameter mapping adjustment is performed on the local area of ​​the current path, and the adjusted feed rate is determined by the ratio between the original set speed and the new and old tool compensation values;

[0034] Rearranging the path in segments based on the adjusted parameters, giving priority to merging adjacent path segments with consistent processing conditions;

[0035] Transition arcs are introduced into the rearranged path structure to replace sharp corner connections, forming a continuous and stable improved path solution.

[0036] Preferably, the optimization of the smoothness of the path inflection points and transition sections includes:

[0037] Identifying the locations of inflection points and direction vectors of adjacent path segments in the improved path solution;

[0038] A local disturbance area is set around the inflection point, and the path points in the area are displaced, with the displacement direction perpendicular to the path forward direction;

[0039] The curvature change rate of the transition section is calculated based on the position of the path point after the disturbance. If the curvature change rate exceeds the set threshold, the disturbance amplitude is reduced and the adjustment is repeated until the inflection point connection tends to a natural transition and the path continuity meets the processing requirements.

[0040] Preferably, verifying the feasibility of the improved path solution includes:

[0041] Importing the improved path plan into a simulation environment to simulate the actual trajectory of the tool moving along the path;

[0042] Detect the contact status between the tool and the workpiece during simulation and mark the areas where overcutting or undercutting occurs;

[0043] The deviation volume of the area where overcutting or undercutting occurs is calculated. If the deviation volume is less than the upper limit of the allowable error, the path is determined to be feasible; otherwise, the path parameters are adjusted and the verification process is repeated.

[0044] On the other hand, the present invention proposes a precision mold design system based on an intelligent optimization algorithm, comprising:

[0045] The pre-processing module is used to collect the 3D geometric data and processing constraints of the mold parts and generate the initial data set;

[0046] A boundary recognition module is used to extract features from the initial data set and identify geometric features and constraint boundaries in the processing path;

[0047] A path generation module is used to construct a multi-objective optimization framework based on the geometric features and constraint boundaries, and generate a set of candidate processing paths through iterative calculation under the multi-objective optimization framework;

[0048] A parameter adjustment module is used to perform multi-dimensional comparative analysis on the candidate processing path set, screen out the Pareto optimal solution set, and dynamically adjust the path parameters based on the Pareto optimal solution set in combination with the process experience database to generate an improved path plan;

[0049] The path optimization verification and output module is used to perform local search on the improved path scheme, optimize the smoothness of the path inflection points and transition sections through the neighborhood perturbation strategy, verify the feasibility of the improved path scheme, output the final processing path and generate CNC code.

[0050] Technical effects and advantages of the present invention: The precision mold design method and system based on intelligent optimization algorithm proposed in the present invention have the following advantages over the existing technology:

[0051] By constructing a multi-objective optimization framework, combining iterative calculations to generate a set of candidate paths, and introducing a Pareto optimal solution screening mechanism, this method can effectively balance multiple key performance indicators, such as path length, cutting load distribution, and surface quality. Furthermore, the method dynamically adjusts path parameters by integrating with a process experience database, and optimizes path inflection points and transition segments through a neighborhood perturbation strategy. This ensures that the resulting machining path not only meets efficiency and stability requirements but also adapts to actual machining needs under different process conditions. This method significantly improves the machining efficiency and surface consistency of complex mold parts, reduces trial-and-error costs, and enables intelligent, adaptive precision mold design and machining path planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flow chart of the precision mold design method based on the intelligent optimization algorithm of the present invention;

[0053] Figure 2 This is a block diagram of the precision mold design system based on the intelligent optimization algorithm of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. 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.

[0055] The present invention provides Figure 1 A precision mold design method based on an intelligent optimization algorithm is shown, comprising the following steps:

[0056] Step 1: Collect the 3D geometric data and processing constraints of the mold parts to generate the initial data set; this includes the following steps:

[0057] Obtain the three-dimensional geometric model of the mold part through laser scanning or CAD modeling system, and extract the curvature distribution information of each surface, including the principal curvature and Gaussian curvature of each point on the model surface, to form a continuous curvature distribution map;

[0058] Discretize the curvature distribution information into multiple curvature intervals (e.g., low curvature, medium curvature, high curvature), and count the point density in each interval;

[0059] Combining the point density and the preset sampling step length, the distribution ratio of the number of sampling points in each area is calculated. The formula is: Sp i =D i / ΣDi , where Sp i represents the sampling ratio of the i-th curvature interval, D i represents the point density within that interval. This formula embodies the "density-driven" sampling philosophy, meaning that areas with denser geometric features (higher point density) should be allocated more sampling resources to ensure accurate representation of key areas. Based on the sampling point allocation ratio, adaptive sampling is performed within each curvature interval to generate a representative spatial point cloud collection.

[0060] Step 2: Extract features from the initial data set to identify geometric features and constraint boundaries in the machining path; including the following steps:

[0061] Perform neighborhood topology analysis on the point cloud set in the initial data set to determine the local surface change trend of each point. Specifically, in the generated spatial point cloud set, construct a local neighborhood area (such as K nearest neighbors or a fixed-radius spherical neighborhood) with each point as the center. By calculating the geometric relationship between the points in the area (such as normal vector consistency, curvature change rate, etc.), the surface change trend of the area where the point is located is determined, whether it is flat, convex or concave.

[0062] Based on the local surface change trend, the curvature mutation area is extracted as the potential geometric feature boundary; specifically, the area with obviously inconsistent local surface change trend (such as the angle between the normal vectors of adjacent points exceeds the set threshold) is marked as the curvature mutation area. These areas usually represent key geometric features such as edges, corners or parting lines on the part.

[0063] By combining the geometric feature boundary with the projection relationship of the machining direction, the infeasible region affected by tool interference is calculated using the formula: I = S × (1-cosθ), where I represents the interference intensity, S represents the contact area, and θ represents the angle between the tool axis and the surface normal. The larger the angle between the tool axis and the surface normal (i.e., the smaller cosθ), the more severe the tool tilt and the higher the risk of interference. Furthermore, if the contact area S is large, the actual interference effect is more significant. Therefore, this formula can be used to quantitatively assess the interference potential of each region. By combining the infeasible region with the geometric feature boundary, the actual usable machining path constraint range is defined.

[0064] Step 3: Constructing a multi-objective optimization framework based on the geometric features and constraint boundaries; including the following steps:

[0065] Based on the geometric feature boundaries and constraint range, the processing area is divided into multiple sub-areas, and the surface inclination angle and contact risk level of each sub-area are marked; the path planning priority is set for each sub-area, and the priority value is determined according to the inverse relationship between the contact risk level and the surface inclination angle. The formula is: Q i =1 / (R i +Ai ), where Q i represents the path planning priority of the i-th sub-area, R i Indicates the exposure risk value of the area, A i It represents the quantitative value of surface inclination; this formula reflects the principle that “low risk + small inclination” areas should be treated first. i Smaller and with an inclination of A i When it is also smaller, its priority Q i The higher the thickness, the more suitable it is for arranging high-quality and efficient cutting paths.

[0066] Combined with the path planning priority, three parallel optimization goals are set: minimizing the idle travel distance, balancing the cutting load distribution, and reducing the number of inflection points; these three goals are not equally important and can be optimized in different sub-areas according to their priority Q i Dynamically adjust weight allocation to ensure that key areas obtain optimal path configuration. Map the priority of each sub-area to the corresponding target weight to form a goal-oriented optimization space structure. i , which is mapped into the weighted coefficients for the above three optimization objectives, for example:

[0067] High priority areas: focus on minimizing idle travel and reducing inflection points;

[0068] Medium priority area: focus on load balancing;

[0069] Low priority area: mainly based on path feasibility.

[0070] Finally, these weighted objectives are integrated into a unified optimization space structure to guide the subsequent path search process.

[0071] Step 4: Generate a set of candidate processing paths through iterative calculation within the multi-objective optimization framework; including the following steps:

[0072] A set of initial path sequences are randomly generated within the optimized spatial structure as the first generation path population; these paths satisfy basic geometric constraints and preliminarily consider the priority distribution of each sub-region.

[0073] The fitness of each path in the initial path population is evaluated. The scoring is based on the weighted sum of path length, load fluctuation amplitude, and number of inflection points. The fitness of each path is evaluated using a weighted scoring mechanism: Score = w1×L+w2×F+w3×T, where:

[0074] L: total length of the path;

[0075] F: cutting load fluctuation value, reflecting the severity of cutting force changes between different areas on the path;

[0076] T: the number of inflection points in the path;

[0077] w1, w2, w3: represent the weight coefficients of the three indicators respectively, the sum of which is 1, and are set according to process requirements.

[0078] This formula embodies the core idea of ​​multi-objective optimization, which is to find a balance between multiple performance indicators. For example, if efficiency is emphasized, w1 is increased; if stability is emphasized, w2 is increased.

[0079] According to the fitness score results, high-scoring paths are selected, and a new batch of paths are generated through cross-joining and mutation operations. The formula is: X new =(X parent1 *W1)+(X parent2 *W2), where X new Represents the position sequence of the newly generated path, X parent1 、(X parent2 represents the position sequence of the parent path, and W1 and W2 represent the weight coefficients of the parent path. This formula simulates the "gene recombination" process in biological genetics, and merges parts of the two paths through linear combination to form a new path with the advantages of both.

[0080] The fitness evaluation and path updating process is repeated until a preset number of iterations (such as 50 or 100 times) is reached, and finally a set of candidate processing paths is output.

[0081] Step 5: Perform multi-dimensional comparative analysis on the candidate processing path set to select the Pareto optimal solution set; including the following steps:

[0082] Each path in the candidate processing path set is ranked based on multi-objective performance, and the independent ranking of each path under the three indicators of path length (the shorter the better), cutting force fluctuation (the smaller the better), and surface quality (the higher the better) is recorded; each path is assigned a ranking value in each indicator (such as 1st, 2nd, etc.) for subsequent comprehensive evaluation.

[0083] Calculate the comprehensive advantage index of each path using the following formula:

[0084] Among them S i represents the comprehensive advantage index of the i-th path, The formula represents the path's ranking in each of the three indicators. The core idea is to sum the rankings of the three indicators and take the inverse, so that paths with higher rankings have higher overall scores. This design avoids subjective bias caused by manually set weights and is suitable for comprehensive evaluation of multiple objectives in the absence of bias.

[0085] Paths are sorted from high to low according to the comprehensive advantage index, and a threshold is set to select the top N items as the preliminary preferred path group; N can be set according to actual needs (such as taking the top 10% or a fixed number such as 10).

[0086] Perform pairwise comparisons within the initial set of preferred paths, excluding completely dominated paths and retaining non-dominated paths to form a Pareto-optimal solution set. This process can be achieved by constructing a dominance matrix or comparing paths one by one, ultimately outputting a set of optimal paths that are mutually non-dominated and have balanced performance.

[0087] Step 6: Based on the Pareto optimal solution set, dynamically adjust the path parameters in combination with the process experience database to generate an improved path plan; including the following steps:

[0088] Match each path in the Pareto optimal solution set with the historical processing parameters in the process experience database to find historical cases with a similarity higher than a set threshold;

[0089] According to the tool compensation value and feed rate setting in the historical case, the parameter mapping adjustment is performed on the local area of ​​the current path. The formula is: V adj =V base *(C new / C old ), where V adj Indicates the adjusted feed rate, V base Indicates the original set speed, C new with C old Represent the new and old tool compensation values ​​respectively; this formula reflects the influence of tool compensation changes on cutting efficiency. When using a tool with a smaller radius (i.e. C new <C old ), in order to ensure cutting stability, the feed speed should be appropriately reduced; vice versa.

[0090] The path is segmented and rearranged based on the adjusted parameters, prioritizing the merging of adjacent segments with consistent machining conditions. This merging operation reduces the number of path switches and mitigates the impact of frequent machine starts and stops. Transition arcs are introduced into the rearranged path structure to replace sharp corners, creating a continuous and stable improved path solution.

[0091] Step 7: Perform a local search on the improved path solution and optimize the smoothness of the path inflection points and transition segments using a neighborhood perturbation strategy; including the following steps:

[0092] Identify the inflection point position and the direction vector of the adjacent path segment in the improved path solution; the inflection point is defined as the point where the change in the direction angle of the path tangent exceeds a certain threshold; the direction vector can be obtained by the coordinate difference between two points, for example, for two points P on the path i (x i ,y i) and P i+1 (x i+1 ,y i+1 ), its direction vector is V=(x i+1 -x i ,y i+1 -y i ).

[0093] A local perturbation zone is defined around the inflection point, and small displacement adjustments are applied to the path points within the zone, with the displacement direction perpendicular to the path's forward direction. The curvature change rate of the transition segment is calculated based on the position of the path points after the perturbation, using the formula: K = (θ2 - θ1) / L, where K represents the curvature change rate, θ2 and θ1 are the tangent angles of the path segments before and after the perturbation, and L represents the length of the perturbation segment. The curvature change rate is a key indicator of path smoothness; a lower K value indicates a smoother transition, which helps minimize tool wear and workpiece surface defects. If the curvature change rate exceeds a set threshold, the perturbation amplitude is reduced and adjustments are repeated until the inflection point connection transitions naturally and the path continuity meets machining requirements.

[0094] Step 8: Verify the feasibility of the improved path solution, output the final processing path and generate NC code; including the following steps:

[0095] Use CAM software or a dedicated simulation system to import the improved path plan into the simulation environment. The system simulates the actual trajectory of the tool along the path based on the path coordinate sequence and the set tool parameters. During the simulation process, the contact status between the tool and the workpiece is detected, and areas where overcutting or undercutting occur are marked. The geometric relationship between the tool and the workpiece model is detected in real time during the simulation run: if the tool cuts into the workpiece beyond the designed contour, it is marked as "overcut"; if the tool does not completely remove the material that should be removed, it is marked as "undercut".

[0096] Calculate the deviation volume for the area where overcut or undercut occurs. The formula is: V err =Σ(h i *A i ), where V err represents the total deviation volume, h i Indicates the height difference of the i-th unit, A i Represents the corresponding area; this formula is based on the idea of ​​the infinitesimal method, which divides the entire error area into multiple small units, calculates the product of their height difference and area respectively, and then accumulates them to obtain a quantitative indicator of the overall error.

[0097] If the deviation volume is less than the upper limit of the allowable error, the path is considered feasible; otherwise, the path parameters are adjusted and the verification process is repeated. A maximum allowable deviation volume threshold Vmax is set as the criterion for whether the path is qualified:

[0098] If V errIf ≤Vmax, the current path is considered to meet the processing requirements and the next step of generating NC code can be entered;

[0099] If V err > Vmax, it is necessary to return to the path optimization stage (such as step six or seven), adjust relevant parameters (such as feed rate and path smoothness), and then re-verify.

[0100] By building a closed-loop verification mechanism, we ensure that the final output path meets engineering precision requirements, significantly reduce trial cutting costs and scrap rates, and improve processing efficiency and finished product quality consistency.

[0101] On the other hand, the present invention proposes a precision mold design system based on an intelligent optimization algorithm, comprising:

[0102] The pre-processing module is used to collect the 3D geometric data and processing constraints of the mold parts and generate the initial data set;

[0103] A boundary recognition module is used to extract features from the initial data set and identify geometric features and constraint boundaries in the processing path;

[0104] A path generation module is used to construct a multi-objective optimization framework based on the geometric features and constraint boundaries, and generate a set of candidate processing paths through iterative calculation under the multi-objective optimization framework;

[0105] A parameter adjustment module is used to perform multi-dimensional comparative analysis on the candidate processing path set, screen out the Pareto optimal solution set, and dynamically adjust the path parameters based on the Pareto optimal solution set in combination with the process experience database to generate an improved path plan;

[0106] The path optimization verification and output module is used to perform local search on the improved path scheme, optimize the smoothness of the path inflection points and transition sections through the neighborhood perturbation strategy, verify the feasibility of the improved path scheme, output the final processing path and generate CNC code.

[0107] In addition, the above modules are also used to implement other steps of the above-mentioned precision mold design method based on intelligent optimization algorithm when executed, as follows:

[0108] Step 1: Generate initial dataset

[0109] Input data: Obtain the 3D model of the mold. The surface curvature distribution shows that there are three main curvature ranges (low, medium, and high).

[0110] Discretization processing: Count the point density of each interval D1=100, D2=200, D3=300, and the total density ΣD j =600.

[0111] Calculate the sampling ratio:

[0112] Sp1=100 / 600=0.1667 (sampling ratio of low curvature area);

[0113] Sp2=200 / 600=0.3333 (middle curvature area);

[0114] Sp3=300 / 600=0.5 (high curvature area);

[0115] Adaptive sampling: Generates point cloud collections proportionally, with more sampling points in high curvature areas (e.g., out of a total of 6,000 points, 3,000 are in high curvature areas).

[0116] Step 2: Identify geometric features and constraint boundaries

[0117] Neighborhood analysis: discovers sudden changes in curvature at corners and marks them as potential feature boundaries.

[0118] Interference calculation: contact area of ​​a certain area S = 50mm 2 , the angle between the tool axis and the normal is θ = 30°, cosθ = 0.866, then the interference intensity I = 50×(1-0.866) = 6.7mm 2 , which is determined to be an infeasible area.

[0119] Constraint range: After excluding the interference area, determine the boundary of the available machining path.

[0120] Step 3: Build a multi-objective optimization framework

[0121] Zoning priority: For a sub-area, the contact risk R = 0.5, and the surface inclination A = 0.2, then Q = 1 / (0.5 + 0.2) = 1.4286.

[0122] Optimization goals: Minimize idle travel (weight 40%), balance load (30%), and reduce inflection points (30%).

[0123] Optimization space: Map the priority Q to the target weight to generate a priority-driven path planning space.

[0124] Step 4: Generate a set of candidate paths

[0125] Initial path: 100 paths are randomly generated (first generation population).

[0126] Fitness evaluation: Path length = 500 mm, load fluctuation = 15%, number of inflection points = 8, weighted score = 500 × 0.4 + 15 × 0.3 + 8 × 0.3 = 200 + 4.5 + 2.4 = 206.9.

[0127] Crossover and mutation: Parent path 1 (X parent1 ) and parent path 2(X parent2) Weights W1 = 0.6, W2 = 0.4, new path X new =X parent1 ×0.6+X parent2 ×0.4.

[0128] Iterative optimization: After 50 iterations, 20 candidate paths are generated.

[0129] Step 5: Screening the Pareto optimal solution set

[0130] Multi-objective ranking: A path ranks first, second, and third in length, cutting force, and quality indicators respectively.

[0131] Comprehensive advantage index: S = 1 / (1+2+3) = 0.1667. After sorting, the top 5 paths are taken as the preferred path group.

[0132] Non-dominated screening: eliminate the three paths that are completely dominated by other paths, and finally retain the two Pareto optimal paths.

[0133] Step 6: Generate an improvement path plan

[0134] History matching: A historical case similarity of >85% is found, and its tool compensation value C old =1.2, C new =1.5.

[0135] Parameter adjustment: original feed speed V base =1000mm / min, after adjustment V adj =1000×(1.5 / 1.2)=1250mm / min.

[0136] Path Rearrangement: Merge adjacent path segments and reduce empty travel by 15%.

[0137] Transition optimization: Insert transition arcs to improve the smoothness of inflection point connections by 30%.

[0138] Step 7: Optimize inflection point smoothness

[0139] Disturbance area: The tangent angles before and after the disturbance at a certain inflection point are θ1 = 45° and θ2 = 60°, and the disturbance section length is L = 5 mm.

[0140] Curvature change rate: K = (60-45) / 5 = 3° / mm. If the threshold is 2° / mm, reduce the disturbance amplitude to L = 7.5 mm and K = 2° / mm.

[0141] Step 8: Verify feasibility

[0142] Simulation detection: Mark 3 overcut areas, height difference h i =0.1mm, area A i =20mm2 (Total 5 units).

[0143] Deviation volume: V err =5×0.1×20=10mm 3 The upper limit is 15mm 3 , determine that the path is feasible.

[0144] Output result: Generate CNC code (G code), improve processing efficiency by 20%, and achieve surface roughness Ra≤1.6μm.

[0145] Summary: Through the above process, the mold processing path reduces idle stroke and improves cutting stability while ensuring accuracy. It also eliminates the need for multiple trial cuts and significantly shortens the development cycle.

[0146] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A precision mold design method based on intelligent optimization algorithm, characterized in that: The following steps are involved: Collect 3D geometric data and processing constraints of mold parts to generate initial data sets; Performing feature extraction on the initial data set to identify geometric features and constraint boundaries in the machining path; Constructing a multi-objective optimization framework based on the geometric features and constraint boundaries, and generating a set of candidate processing paths through iterative calculation within the multi-objective optimization framework; Performing a multi-dimensional comparative analysis on the candidate processing path set to screen out a Pareto optimal solution set, dynamically adjusting path parameters based on the Pareto optimal solution set and combining with a process experience database to generate an improved path solution; A local search is performed on the improved path scheme, and the smoothness of the path inflection points and transition sections is optimized through a neighborhood perturbation strategy to verify the feasibility of the improved path scheme, output the final processing path, and generate the NC code.

2. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: The method comprises the following steps, including: Obtain the three-dimensional geometric model of the mold part and extract the curvature distribution information of each surface; Discretize the curvature distribution information into multiple curvature intervals, and count the point density in each interval; Combining the point density with the preset sampling step length, calculating the distribution ratio of the number of sampling points in each area; According to the sampling point quantity distribution ratio, adaptive sampling is performed in each curvature interval to generate a spatial point cloud set.

3. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: The identification of geometric features and constraint boundaries in the machining path includes: Performing neighborhood topology analysis on the point cloud set in the initial data set to determine the local surface change trend of each point; According to the local surface change trend, extracting the curvature mutation area as the potential geometric feature boundary; Combining the geometric feature boundary with the machining direction projection relationship to calculate the infeasible area affected by tool interference; The infeasible area and the geometric feature boundary are combined to define the actual available machining path constraint range.

4. The precision mold design method based on intelligent optimization algorithm according to claim 3, characterized in that: A multi-objective optimization framework is constructed based on the geometric features and constraint boundaries, including: Dividing the processing area into multiple sub-areas based on the geometric feature boundaries and constraint range, and marking the surface inclination angle and contact risk level of each sub-area; A path planning priority is set for each sub-area. Based on the path planning priority, three parallel optimization objectives are set: minimizing the idle travel distance, balancing the cutting load distribution, and reducing the number of inflection points. The priority of each sub-region is mapped to the corresponding target weight to form a target-oriented optimization spatial structure.

5. The precision mold design method based on intelligent optimization algorithm according to claim 4, characterized in that: The generating of a candidate processing path set by iterative calculation includes: Randomly generating a set of initial path sequences within the optimized spatial structure as a first generation path population; Performing fitness evaluation on each path in the initial path population, wherein the scoring is based on a weighted sum of path length, load fluctuation amplitude, and number of inflection points; Based on the fitness score results, high-scoring paths are selected and a new batch of paths are generated through cross-joining and mutation operations. The position sequence of the new path is generated by a weighted combination of the position sequence of the parent path and the weight coefficient; The fitness evaluation and path updating process is repeated until the preset number of iterations is reached, and finally a set of candidate processing paths is output.

6. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: The screening out of the Pareto optimal solution set includes: Performing multi-objective performance ranking on each path in the candidate machining path set, and recording the independent ranking of each path under three indicators: path length, cutting force fluctuation, and surface quality; Calculate the comprehensive advantage index of each path; Sort the paths from high to low according to the comprehensive advantage index, and set a threshold to select the top N paths as the preliminary preferred path group; Pairwise comparisons are performed within the preliminary preferred path group, completely dominated paths are excluded, and non-dominated paths are retained to form a Pareto optimal solution set.

7. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: The generating of the improvement path plan includes: Matching each path in the Pareto optimal solution set with historical processing parameters in a process experience database to find historical cases with a similarity higher than a set threshold; Based on the tool compensation value and feed rate settings in the historical case, parameter mapping adjustment is performed on the local area of ​​the current path, and the adjusted feed rate is determined by the ratio between the original set speed and the new and old tool compensation values; Rearranging the path in segments based on the adjusted parameters, giving priority to merging adjacent path segments with consistent processing conditions; Transition arcs are introduced into the rearranged path structure to replace sharp corner connections, forming a continuous and stable improved path solution.

8. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: The optimization of the smoothness of the path inflection points and transition sections includes: Identifying the locations of inflection points and direction vectors of adjacent path segments in the improved path solution; A local disturbance area is set around the inflection point, and the path points in the area are displaced, with the displacement direction perpendicular to the path forward direction; The curvature change rate of the transition section is calculated based on the position of the path point after the disturbance. If the curvature change rate exceeds the set threshold, the disturbance amplitude is reduced and the adjustment is repeated until the inflection point connection tends to a natural transition and the path continuity meets the processing requirements.

9. The precision mold design method based on intelligent optimization algorithm according to claim 1, characterized in that: Verify the feasibility of the improvement path plan, including: Importing the improved path plan into a simulation environment to simulate the actual trajectory of the tool moving along the path; Detect the contact status between the tool and the workpiece during simulation and mark the areas where overcutting or undercutting occurs; The deviation volume of the area where overcutting or undercutting occurs is calculated. If the deviation volume is less than the upper limit of the allowable error, the path is determined to be feasible; otherwise, the path parameters are adjusted and the verification process is repeated.

10. A precision mold design system based on an intelligent optimization algorithm for implementing the method according to any one of claims 1 to 9, characterized in that: include: The pre-processing module is used to collect the 3D geometric data and processing constraints of the mold parts and generate the initial data set; A boundary recognition module is used to extract features from the initial data set and identify geometric features and constraint boundaries in the processing path; A path generation module is used to construct a multi-objective optimization framework based on the geometric features and constraint boundaries, and generate a set of candidate processing paths through iterative calculation under the multi-objective optimization framework; A parameter adjustment module is used to perform multi-dimensional comparative analysis on the candidate processing path set, screen out the Pareto optimal solution set, and dynamically adjust the path parameters based on the Pareto optimal solution set in combination with the process experience database to generate an improved path plan; The path optimization verification and output module is used to perform local search on the improved path scheme, optimize the smoothness of the path inflection points and transition sections through the neighborhood perturbation strategy, verify the feasibility of the improved path scheme, output the final processing path and generate CNC code.

Images