Result analysis method for numerical control machining geometric simulation
Through topological structure and spatial mesh division methods, the problems of low efficiency and low accuracy in the analysis of CNC machining simulation results are solved, and efficient and accurate error analysis and visual result display are achieved.
Patent Information
- Application Number
- CN202510632745.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-16
- Publication Date
- 2025-08-08
AI Technical Summary
The existing CNC machining simulation results analysis algorithm is low efficiency and low accuracy, especially the Boolean operation method based on solid model is low efficiency, and the error results of the method based on surface discrete are largely different from the actual results.
The region diffusion and spatial meshing method based on topology is adopted, and error analysis is achieved by calculating the sampling points of the simulation result model and the directed distance of the design model, combining spatial meshing and sampled point cloud rendering.
The calculation efficiency and accuracy of CNC machining simulation results analysis can be improved, over-cut and under-cut areas can be clearly identified, the sampling point generation algorithm is optimized, and the algorithm accuracy and calculation efficiency are balanced.
Smart Images
Figure CN120449495A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of numerical control machining simulation, and in particular relates to a result analysis method for numerical control machining geometric simulation. Background Art
[0002] To determine whether the machining program is correct, after the NC machining geometry simulation is completed, it is necessary to compare the error between the simulation result model and the design model to verify whether the machining results meet the accuracy requirements. Simulation result analysis algorithms can be divided into analysis algorithms based on solid models and analysis algorithms based on surface discretization.
[0003] Analysis algorithms based on solid models are usually implemented through Boolean operations on solid models. Assume that model A is the simulation result model and model B is the design model. The result of the Boolean operation (AB)∪(BA) is the error analysis result of models A and B. This method is inefficient and not very practical. Cignoni et al. first proposed an error analysis algorithm for STL models. The error value was obtained by sampling the triangular faces of one model and calculating the Hausdorff distance between the sampling points and the other model. Kang et al. proposed a bilateral Hausdorff distance method for calculating the triangular mesh model to the quadrilateral mesh model, and used it to evaluate the error between the two models. Miao used the spatial grid method to reduce the calculation amount of the Hausdorff distance and improve the efficiency of the error analysis algorithm.
[0004] Through the surface discretization analysis algorithm, the discrete errors in all directions on the surface can be approximated to the actual shape and position errors. Among them, the Z-direction error method is the most common surface discretization analysis method. It mainly uses the numerical value of the Z-direction vector to represent the height value of the model under the premise of Boolean operation simulation of discrete vector intersection to directly compare the Z-direction error value to achieve error analysis. However, the error results of this method are quite different from the actual results. Based on the subdivision of the blank, Wu Tiejun proposed a local subdivision error analysis (Z&N) algorithm. The verification accuracy of this method is greater than that of the Z-direction error method. Zhao Jin proposed a normal error verification method based on triangulation discrete simulation, which overcomes the disadvantage of the Z&N algorithm that the error will be amplified in steep areas of the surface. Summary of the Invention
[0005] In response to the above problems, the present invention provides a result analysis method for CNC machining geometric simulation.
[0006] A result analysis method for numerical control machining geometry simulation of the present invention comprises the following steps:
[0007] Step 1: Calculation of sampling points on the surface of the simulation result model.
[0008] Calculate the sampling points on each triangular face of the simulation result model, use the two sides of the triangular face as the generation direction of the sampling point array, and the number of sampling points on each side is equal.
[0009] On triangle A, along vector P A_0 P A_1 and P A_0 P A_2 Generate a sampling point matrix in the direction of , with 6 sampling points on each side; A_0 P A_1 The i-th vector P in the direction A_0 P A_2 The jth sampling point in the direction is defined as sampling point D A_i_j .
[0010] Let the sampling density of the triangle be 1 / s 2 , where s is the set sampling step size; if the number of sampling points on each edge of the triangle is n e , then the total number of sampling points of the triangle is n e (n e +1) / 2; the relationship between the sampling density of the triangle and the number of sampling points is shown in formula (1):
[0011]
[0012] Where S T The area of the triangle is given by the following formula:
[0013]
[0014] If n is calculated by formula (2) e The value is not an integer and needs to be rounded to determine the final number of sampling points on each edge of the triangle. The calculation formula of the sampling point coordinates is as follows:
[0015] D A_i_j =P A_0 +ie1+je2,0≤i≤n e -1, 0≤j≤n e -1-i(3)
[0016] Where:
[0017]
[0018] Step 2: Space grid division.
[0019] First, calculate the union of the AABB bounding boxes of the simulation result model and the design model, and set the coordinates of the minimum point as (x min ,y min ,z min), the maximum point coordinate is (x max ,y max ,z max ); then, the bounding box union is spatially partitioned using cubic cells of equal volume, where the cell size is controlled by the side length Δd; the value of Δd is determined as follows: an equilateral triangle is constructed whose area is equal to the average area of all triangular faces of the design model, and the value of Δd is set to the integer part of the side length of the equilateral triangle.
[0020] Define the number of spatial cells in the X direction as N x , which is calculated as follows:
[0021]
[0022] Similarly, the number of spatial cells N in the Y and Z directions is also calculated. y and N z ; Define the number of the spatial cell as C(x,y,z), where x∈[0,N x ),y∈[0,N y ),z∈[0,N z ); the total number of spatial cells is N x ×N y ×N z .
[0023] Based on the separating axis theorem, the intersection of the triangular faces of the design model and the spatial cells is determined, and the index values of the intersecting triangular faces are recorded in the corresponding numbered spatial cells.
[0024] Step 3: Calculation of sampling point error.
[0025] The error calculation of the sampling point includes three steps: calculating the spatial cell number where the sampling point is located, calculating the initial value of the directed distance from the sampling point to the design model, and calculating the final directed distance value from the sampling point to the design model.
[0026] (1) Determine the spatial cell number where the sampling point is located.
[0027] Given the coordinates of a sampling point, quickly locate the spatial cell number C(x,y,z) where it is located by bisection; at the same time, define L n (C) is a set of cells that are n*Δd away from cell C.
[0028] (2) Calculate the initial value of the directed distance.
[0029] Define the directed distance from the sampling point D to the design model S as d(D, S), and set the initial value of d(D, S) to +∞; traverse all the triangular faces intersecting with the cell C(x, y, z), and calculate the directed distance d(D, T) from the sampling point D to the triangular face T; the calculation method of d(D, T) is as follows: Solve the projection point D' of the sampling point D on the plane where the triangular face T is located, and determine whether D' is inside the triangular face.
[0030] If D' is inside the triangular face T, the calculation formula of |d(D, T)| is as follows:
[0031] |d(D, T)| = ||D - D'||2 (6)
[0032] If D' is not inside the triangular face T, solve the point D” on the three sides of the triangular face T that is closest to the projection point D'; at this time, the calculation formula of |d(D, T)| is as follows:
[0033] |d(D, T)| = ||D - D”||2 (7)
[0034] Let the normal vector of the triangular face T be N, then the sign of the directed distance is the same as the sign of the formula N(D - D'); if the calculated |d(D, T)| is less than |d(D, S)|, then update d(D, S) = d(D, T).
[0035] (3) Calculate the final directed distance value.
[0036] Calculate the directed distances from the sampling point D to all the triangular faces intersecting with the cell set L1(C), and update the value of d(D, S); if |d(D, S)| < n * Δd, then the value of d(D, S) is the final directed distance from the sampling point D to the design model S; otherwise, increase the value of n in turn, and calculate the directed distances from the sampling point D to all the triangular faces intersecting with the cell set L n (C), and update the value of d(D, S), until the condition |d(D, S)| < n * Δd is satisfied.
[0037] The sampling point D is located in the cell C, and the initial value of the directed distance d(D, S) calculated in C is dis1; traverse the first-layer cell set L1(C) around C, the minimum distance from the sampling point to the triangular faces intersecting with the cells in L1(C) is dis2, and since dis2 < dis1, the value of d(D, S) is updated to dis2; in addition, since |d(D, S)| < 1 * Δd, there is no need to traverse the second-layer cell set L2(C) around the cell C; the current value of d(D, S) is the final directed distance value from the sampling point to the design model.
[0038] Step 4: Visualize the analysis results.
[0039] Define the overcutting accuracy value ε over and undercut accuracy value ε under , both are non-negative numbers; according to the calculated sampling point error value, it is divided into three types. Taking the sampling point D as an example, if d(D,S)∈[-ε over ,ε under ], then the sampling point D is an error-free point; if d(D,S)∈[-∞,-ε over ), then the sampling point D is the over-cut point; if d(D,S)∈(ε under ,+∞], then the sampling point D is an undercut point.
[0040] When displaying the cloud map, the simulation result model is represented as a sampling point cloud. The color of each sampling point is determined according to the sampling point type divided above. The change of color represents different error areas. The blue point cloud represents the under-cut area, the red point cloud represents the over-cut area, and the green point cloud represents the part that meets the processing accuracy requirements.
[0041] Furthermore, in step 1, the number of redundant sampling points can be reduced by simulating the topological structure of the result model, thereby optimizing the sampling point generation algorithm. The specific optimization steps are as follows:
[0042] S1: Construct the topological structure of the simulation result model, select any triangular face in the model as the initial seed face, calculate the sampling points of the triangular face, and store the sampling points inside the triangular face into the sampling point set.
[0043] S2: Find the adjacent triangles of the initial seed face and calculate the sampling points of the adjacent triangles. For the edge shared by two triangles, select the side with more sampling points as the sampling points of the edge, and store the sampling points inside the adjacent triangles and the sampling points of the shared edge into the sampling point set.
[0044] S3: Continue to diffuse adjacent triangles and obtain sampling points of the triangles until all triangles of the simulation result model are traversed.
[0045] S4: Store the triangle vertices in the simulation result model vertex table into the sampling point set.
[0046] S5: Sampling ends.
[0047] Furthermore, in step 3, in order to balance the accuracy and computational efficiency of the algorithm, the sampling step s is set to 0.2% of the diagonal length of the bounding box of the simulation result model.
[0048] The beneficial technical effects of the present invention are:
[0049] Aiming at the analysis and research of CNC machining simulation results, the present invention adopts the regional diffusion and spatial grid division method based on topological structure, improves the simulation result sampling point calculation and error calculation method, and improves the computational efficiency of the result analysis algorithm. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 Schematic diagram of triangle sampling points.
[0051] Figure 2 Schematic diagram of the final directed distance calculation from the sampling point to the design model.
[0052] Figure 3 is the relationship between sampling step size and computational efficiency.
[0053] Figure 4 It is the classification of sampling point error levels.
[0054] Figure 5 Schematic diagram of the relationship between the simulation result model and the design model.
[0055] Figure 6 A schematic diagram showing the sampling point cloud. DETAILED DESCRIPTION
[0056] The present invention is further described in detail below with reference to the accompanying drawings and specific examples.
[0057] A result analysis method for numerical control machining geometry simulation of the present invention comprises the following steps:
[0058] Step 1: Calculation of sampling points on the surface of the simulation result model.
[0059] To represent the simulation result model as a point cloud, it is necessary to calculate the sampling points on each triangular face. This paper adopts a specific triangular face sampling point calculation method. The characteristic of this method is that the two sides of the triangular face are used as the generation direction of the sampling point array, and the number of sampling points on each side is equal.
[0060] On triangle A, along vector P A_0 P A_1 and P A_0 P A_2 Generate a sampling point matrix in the direction of , with 6 sampling points on each side; A_0 P A_1 The i-th vector P in the direction A_0 P A_2 The jth sampling point in the direction is defined as sampling point D A_i_j ,like Figure 1 shown.
[0061] Let the sampling density of the triangle be 1 / s 2, where s is the set sampling step size; if the number of sampling points on each edge of the triangle is n e , then the total number of sampling points of the triangle is n e (n e +1) / 2; the relationship between the sampling density of the triangle and the number of sampling points is shown in formula (1):
[0062]
[0063] Where S T The area of the triangle is given by the following formula:
[0064]
[0065] If n is calculated by formula (2) e The value is not an integer and needs to be rounded to determine the final number of sampling points on each edge of the triangle. The calculation formula of the sampling point coordinates is as follows:
[0066] D A_i_j =P A_0 +ie1+je2,0≤i≤n e -1, 0≤j≤n e -1-i(3)
[0067] Where:
[0068]
[0069] Since each edge in the simulation result model is shared by two triangles, and each vertex is shared by more than two triangles, when sampling this model, if the sampling points of each triangle are calculated separately, a large number of redundant sampling points will be generated at the vertices and edges, thereby reducing the computational efficiency of the algorithm. By using the topological structure of the simulation result model, the number of redundant sampling points can be effectively reduced, thereby optimizing the sampling point generation algorithm. The specific optimization steps are as follows:
[0070] S1: Construct the topological structure of the simulation result model, select any triangular face in the model as the initial seed face, calculate the sampling points of the triangular face, and store the sampling points inside the triangular face into the sampling point set.
[0071] S2: Find the adjacent triangles of the initial seed face and calculate the sampling points of the adjacent triangles. For the edge shared by two triangles, select the side with more sampling points as the sampling points of the edge, and store the sampling points inside the adjacent triangles and the sampling points of the shared edge into the sampling point set.
[0072] S3: Continue to diffuse adjacent triangles and obtain sampling points of the triangles until all triangles of the simulation result model are traversed.
[0073] S4: Store the triangle vertices in the simulation result model vertex table into the sampling point set.
[0074] S5: Sampling ends.
[0075] Step 2: Space grid division.
[0076] First, calculate the union of the AABB bounding boxes of the simulation result model and the design model, and set the coordinates of the minimum point as (x min ,y min ,z min ), the maximum point coordinate is (x max ,y max ,z max ); then, the bounding box union is spatially partitioned using cubic cells of equal volume, where the cell size is controlled by the side length Δd; the value of Δd is determined as follows: an equilateral triangle is constructed whose area is equal to the average area of all triangular faces of the design model, and the value of Δd is set to the integer part of the side length of the equilateral triangle.
[0077] Define the number of spatial cells in the X direction as N x , which is calculated as follows:
[0078]
[0079] Similarly, the number of spatial cells N in the Y and Z directions is also calculated. y and N z ; Define the number of the spatial cell as C(x,y,z), where x∈[0,N x ),y∈[0,N y ),z∈[0,N z ); the total number of spatial cells is N x ×N y ×N z .
[0080] Based on the separating axis theorem, the intersection of the triangular faces of the design model and the spatial cells is determined, and the index values of the intersecting triangular faces are recorded in the corresponding numbered spatial cells.
[0081] Step 3: Calculation of sampling point error.
[0082] The error calculation of the sampling point includes three steps: calculating the spatial cell number where the sampling point is located, calculating the initial value of the directed distance from the sampling point to the design model, and calculating the final directed distance value from the sampling point to the design model.
[0083] (1) Determine the spatial cell number where the sampling point is located.
[0084] Given the coordinates of a sampling point, quickly locate the spatial cell number C(x,y,z) where it is located by bisection; at the same time, define L n (C) is a set of cells that are n*Δd away from cell C.
[0085] (2) Calculate the initial value of the directed distance.
[0086] Define the directed distance from the sampling point D to the design model S as d(D, S), and set the initial value of d(D, S) to +∞; traverse all triangles that intersect with the cell C(x, y, z) and calculate the directed distance d(D, T) from the sampling point D to the triangle T; the calculation method of d(D, T) is as follows: solve the projection point D' of the sampling point D to the plane where the triangle T is located, and determine whether D' is inside the triangle.
[0087] If D' is inside the triangle T, the calculation formula for |d(D,T)| is as follows:
[0088] |d(D,T)|=||D-D'||2 (6)
[0089] If D' is not inside the triangle T, find the point D'' on the three edges of the triangle T that is closest to the projected point D'. In this case, the calculation formula for |d(D,T)| is as follows:
[0090] |d(D,T)|=||DD”||2 (7)
[0091] Assume that the normal vector of triangle T is N, then the sign of the signed distance is consistent with the sign of the formula N(D-D'); if the calculated |d(D,T)| is less than |d(D,S)|, then update d(D,S)=d(D,T).
[0092] (3) Calculate the final directed distance value.
[0093] Calculate the directed distance from the sampling point D to all triangles intersecting with the cell set L1(C), and update the value of d(D, S); if |d(D, S)| < n*Δd, then the value of d(D, S) is the final directed distance from the sampling point D to the design model S; otherwise, increase the value of n in sequence, and calculate the distance from the sampling point D to the triangles intersecting with the cell set L1(C). n (C) The directed distances of all intersecting triangles are calculated and the value of d(D, S) is updated until the condition |d(D, S)| < n*Δd is satisfied.
[0094] like Figure 2As shown in the figure, the sampling point D is located in the cell C, and the initial value of the directed distance d(D, S) calculated in C is dis1; traverse the first-layer cell set L1(C) around C, and the minimum distance from the sampling point to the triangular faces intersecting with the cells in L1(C) is dis2. Since dis2 < dis1, the value of d(D, S) is updated to dis2; in addition, since |d(D, S)| < 1 * Δd, there is no need to traverse the second-layer cell set L2(C) around the cell C; the current value of d(D, S) is the final directed distance value from the sampling point to the design model.
[0095] According to Equation (2), it can be seen that the smaller the sampling step size s, the more sampling points, the more accurate the point cloud representation of the simulation result model, and the more accurate the calculation result of the error. However, the calculation efficiency of the algorithm will decrease accordingly. Therefore, in order to balance the accuracy and calculation efficiency of the algorithm, a suitable sampling step size needs to be set. The algorithm development environment provided by this invention is shown in Table 1.
[0096] Table 1 Algorithm development environment
[0097]
[0098] Under the development environment, complete the development of the simulation result analysis algorithm, and draw a line chart of the calculation efficiency and sampling step size of three different specification result model analysis cases, as Figure 3 shown. The test results show that when the sampling step size s is set to 0.2% of the diagonal length of the bounding box of the simulation result model, it can better balance the accuracy and calculation efficiency of the algorithm.
[0099] Step 4: Visualize the analysis results.
[0100] In order to visualize the over-cut area and under-cut area of the result analysis, define the over-cut precision value ε over and the under-cut precision value ε under , both of which are non-negative numbers; according to the calculated sampling point error values, divide them into three types. Taking the sampling point D as an example, if d(D, S) ∈ [-ε over , ε under , then the sampling point D is a point without error; if d(D, S) ∈ [-∞, -ε over ), then the sampling point D is an over-cut point; if d(D, S) ∈ (ε under , +∞], then the sampling point D is an under-cut point, as Figure 4 shown.
[0101] When displaying in a cloud map, represent the simulation result model as a sampling point cloud, and the color of each sampling point is determined according to the above-mentioned divided sampling point types. The change in color represents different error regions. The blue point cloud represents the under-cut region, the red point cloud represents the over-cut region, and the green point cloud represents the part that meets the machining accuracy requirements.
[0102] Algorithm verification:
[0103] The sampling optimization of the topological structure and the efficiency optimization method of the spatial grid were tested, as shown in Table 2. The test results show that the sampling based on the topological structure and the division of the spatial grid can effectively improve the efficiency of the simulation result analysis algorithm.
[0104] Table 2 Algorithm test results
[0105]
[0106] Two surface models are established by UG software, one of which is used as the simulation result model and the other as the design model. Figure 5 The simulation result model is analyzed for errors, and the analysis results are rendered through the sampling point cloud, as shown in Figure 6 The test results show that the spatial distribution characteristics of overcut and undercut areas can be clearly identified through point cloud visualization technology.
Claims
1. A method for analyzing results of numerical control machining geometry simulation, characterized in that: It includes the following steps: Step 1: Calculation of sampling points on the surface of the simulation result model; Calculate the sampling points on each triangular face of the simulation result model. Take the two sides of the triangular face as the generation directions of the sampling dot matrix, and the number of sampling points on each side is equal; On triangle A, along vector P A_0 P A_1 and P A_0 P A_2 Generate a sampling point matrix in the direction of , with 6 sampling points on each side; A_0 P A_1 The i-th vector P in the direction A_0 P A_2 The jth sampling point in the direction is defined as sampling point D A_i_j ; Let the sampling density of the triangle be 1 / s 2 , where s is the set sampling step size; if the number of sampling points on each edge of the triangle is n e , then the total number of sampling points of the triangle is n e (n e +1) / 2; the relationship between the sampling density of the triangle and the number of sampling points is shown in formula (1): Where S T The area of the triangle is given by the following formula: If n is calculated by formula (2) e The value is not an integer and needs to be rounded to determine the final number of sampling points on each edge of the triangle. The calculation formula of the sampling point coordinates is as follows: D A_i_j =P A_0 +ie1+je2,0≤i≤n e -1, 0≤j≤n e -1-i(3) where: Step 2: Spatial grid division; First, calculate the union of the AABB bounding boxes of the simulation result model and the design model, and set the coordinates of the minimum point as (x min ,y min ,z min ), the maximum point coordinate is (x max ,y max ,z max ); then, the bounding box union is spatially partitioned using cubic cells of equal volume, where the cell size is controlled by the side length Δd; the value of Δd is determined as follows: construct an equilateral triangle whose area is equal to the average area of all triangular faces of the design model, and set the value of Δd to the integer part of the side length of the equilateral triangle; Define the number of spatial cells in the X direction as N x , which is calculated as follows: Similarly, the number of spatial cells N in the Y and Z directions is also calculated. y and N z ; Define the number of the spatial cell as C(x,y,z), where x∈[0,N x ),y∈[0,N y ),z∈[0,N z ); the total number of spatial cells is N x ×N y ×N z ; Based on the separating axis theorem, judge the intersection situation between the triangular faces of the design model and the spatial cells, and record the index values of the triangular faces intersecting with them in the corresponding numbered spatial cells; Step 3: Calculation of sampling point error; The error calculation of the sampling points includes three steps: calculating the spatial cell number where the sampling point is located, calculating the initial value of the directed distance from the sampling point to the design model, and calculating the final directed distance value from the sampling point to the design model; (1) Determine the spatial cell number where the sampling point is located; Given the coordinates of a sampling point, quickly locate the spatial cell number C(x,y,z) where it is located by bisection; at the same time, define L n (C) is the set of cells that are n*Δd away from cell C; (2) Calculate the initial value of the directed distance; Define the directed distance from the sampling point D to the design model S as d(D, S), and set the initial value of d(D, S) to +∞; traverse all triangular faces intersecting with the cell C(x, y, z), and calculate the directed distance d(D, T) from the sampling point D to the triangular face T; the calculation method of d(D, T) is as follows: Solve the projection point D' of the sampling point D on the plane where the triangular face T is located, and judge whether D' is located inside the triangular face; If D' is located inside the triangular face T, the calculation formula of |d(D, T)| is as follows: |d(D, T)| = ||D - D'||2 (6) If D' is not located inside the triangular face T, solve the point D” on the three sides of the triangular face T that is closest to the projection point D', at this time, the calculation formula of |d(D, T)| is as follows: |d(D, T)| = ||D - D”||2 (7) Let the normal vector of the triangular face T be N, then the positive and negative of the directed distance are the same as the positive and negative of the formula N(D - D'); if the calculated |d(D, T)| is less than |d(D, S)|, then update d(D, S) = d(D, T); (3) Calculate the final directed distance value; Calculate the directed distance from the sampling point D to all triangles intersecting with the cell set L1(C), and update the value of d(D, S); if |d(D, S)| < n*Δd, then the value of d(D, S) is the final directed distance from the sampling point D to the design model S; otherwise, increase the value of n in sequence, and calculate the distance from the sampling point D to the triangles intersecting with the cell set L1(C). n (C) The directed distances of all intersecting triangles are calculated, and the value of d(D, S) is updated until the condition |d(D, S)| < n*Δd is satisfied; The sampling point D is located in the cell C, and the initial value of the directed distance d(D, S) calculated in C is dis1; traverse the first-layer cell set L1(C) around C, the minimum distance from the sampling point to the triangular faces intersecting with the cells in L1(C) is dis2, and since dis2 < dis1, the value of d(D, S) is updated to dis2; in addition, since |d(D, S)| < 1 * Δd, it is not necessary to traverse the second-layer cell set L2(C) around the cell C; the current value of d(D, S) is the final directed distance value from the sampling point to the design model; Step 4: Visualization of analysis results; Define the overcutting accuracy value ε over and undercut accuracy value ε under , both are non-negative numbers; according to the calculated sampling point error value, it is divided into three types. Taking the sampling point D as an example, if d(D,S)∈[-ε over ,ε under ], then the sampling point D is an error-free point; if d(D,S)∈[-∞,-ε over ), then the sampling point D is the over-cut point; if d(D,S)∈(ε under ,+∞], then the sampling point D is an undercut point; When displaying in a contour map, represent the simulation result model as a sampling point cloud. The color of each sampling point is determined according to the above-mentioned divided sampling point types, and the change of color represents different error regions. The blue point cloud represents the undercut region, the red point cloud represents the overcut region, and the green point cloud represents the part that meets the machining accuracy requirements.
2. The result analysis method for CNC machining geometric simulation according to claim 1, characterized in that: In the above Step 1, through the topological structure of the simulation result model, the number of redundant sampling points can be reduced, and then the sampling point generation algorithm can be optimized. The specific optimization steps are as follows: S1: Construct the topology of the simulation result model, select any triangular face in the model as the initial seed face, calculate the sampling points of the triangular face, and store the sampling points inside the triangular face into the sampling point set; S2: Find the adjacent triangles of the initial seed face and calculate the sampling points of the adjacent triangles. For the edge shared by two triangles, select the side with more sampling points as the sampling points of the edge. Store the sampling points inside the adjacent triangles and the sampling points of the shared edge into the sampling point set. S3: Continue to diffuse adjacent triangles and obtain sampling points of the triangles until all triangles of the simulation result model are traversed; S4: storing the triangle vertices in the vertex table of the simulation result model into the sampling point set; S5: Sampling ends.
3. The result analysis method for CNC machining geometric simulation according to claim 1, characterized in that: In step 3, in order to balance the accuracy and computational efficiency of the algorithm, the sampling step length s is set to 0.2% of the diagonal length of the bounding box of the simulation result model.