A machine vision-based simulation virtual processing system

Abstract

This invention discloses a machine vision-based simulation virtual machining system, belonging to the field of machine vision technology. The system includes a model import and preprocessing module, a machine vision recognition and calibration module, a machining path planning module, and a virtual simulation and optimization module. Its key technical points are: by adopting a comprehensive simulation testing and path optimization technology scheme, closed-loop verification of the machining path is achieved, thus preventing potential risks. A highly realistic virtual machining environment is constructed to perform comprehensive simulation testing on each path, checking for overcutting or collision phenomena. For problematic path segments, the system finds alternative path segments for replacement and performs a secondary comparison with the original path after replacement, thereby selecting the optimal path. This reduces the amount of secondary optimization processing required, improving machining safety and stability while also enhancing work efficiency.

Description

Technical Field

[0001] This invention relates to the field of machine vision technology, specifically to a machine vision-based simulation virtual processing system. Background Technology

[0002] Machine vision aims to enable computers or machines to see and understand their surroundings like humans. Specifically, machine vision systems capture images of target objects using cameras or other image acquisition devices, and then use complex algorithms to process and analyze these images to extract useful information or make decisions. Machine vision can be used in a variety of applications, including but not limited to: quality inspection, measurement, recognition and tracking, robot guidance, and virtual manufacturing simulation.

[0003] In the traditional field of simulation-based virtual machining, several significant problems exist in CNC machining path planning and optimization. Firstly, regarding precise workpiece positioning and dynamic adjustment, traditional methods rely on manual calibration, which is not only inefficient but also prone to human error. For example, in machining complex curved surface parts, technicians need to spend a significant amount of time on measurement and calibration, and even then, high-precision positioning is difficult to guarantee, potentially leading to a final product that does not meet design requirements. Secondly, in the path planning stage, traditional techniques often struggle to balance machining time and material consumption. Typically, path planning relies on pre-programmed paths, which, while achieving machining efficiency, fail to provide a more effective balance for other aspects, such as machining accuracy. Thirdly, traditional path planning methods often employ a global replacement strategy, assuming the path is an indivisible whole, and replacing the entire path is the mainstream approach. This leads to local defects potentially rendering the entire path obsolete, increasing the cost and time of replanning. For instance, when machining large mold inserts, if a section of the path has a problem, the entire path needs to be redesigned or replaced with another path that requires verification, easily resulting in significant resource waste and extended production cycles. Summary of the Invention

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A machine vision-based simulation virtual processing system includes:

[0006] The model import and preprocessing module imports the 3D model of the target workpiece, performs format conversion, detects potential problem sources, compares it with the repair solutions given in the solution library, completes the repair process, and completes data standardization.

[0007] The machine vision recognition and calibration module uses machine vision technology to locate and capture the target workpiece and match its features. When a deviation is detected, it triggers an adjustment and recalibration mechanism to ensure that the target workpiece meets the design requirements.

[0008] The machining path planning module initially sets several tool paths based on the shape of the target workpiece and the machining objectives. It then performs a simulation evaluation mechanism on different tool path schemes to filter out candidate paths and triggers a second simulation evaluation mechanism to extract the qualified path from the candidate paths.

[0009] The virtual simulation and optimization module enters the simulation stage after completing path planning, executes the simulation test strategy, tests the processing target, and decides whether to trigger the path optimization mechanism based on the test results to generate the final path; and performs secondary optimization on the final path to generate the optimal path.

[0010] Furthermore, during format conversion, the supported file formats include at least STL and STEP; when detecting potential problem sources, the potential problem sources include at least non-manifold edges, self-intersecting surfaces, and topological errors; the corresponding repair solutions provided by the benchmark solution library include at least non-manifold edge repair solutions, self-intersecting surface repair solutions, and geometric topology repair solutions; data standardization processing: the imported 3D model is standardized in terms of size units.

[0011] Furthermore, the process of locating and capturing the target workpiece is as follows: by using a pre-installed vision sensor network to capture the surface image of the target workpiece in real time, and using an edge detection algorithm to identify contour features; the process of feature matching is as follows: using a feature point matching algorithm to compare the contour features with the corresponding features in the 3D model to complete the positioning operation.

[0012] Furthermore, the triggered adjustment and recalibration mechanism process is as follows: when a deviation is detected, the preset processing parameters are adjusted, and a calibration notification is issued to prompt a second calibration action to complete the recalibration or second adjustment.

[0013] Furthermore, the initial path setting is based on the target workpiece geometry and machining target. By identifying boundary contours, feature regions and machining depth, several tool paths are generated.

[0014] The machining strategies employed include at least: parallel cutting, circumferential cutting, and reciprocating paths.

[0015] Furthermore, the process of executing a simulation evaluation mechanism is as follows:

[0016] Given a set of paths P = {P1, P2, ..., P...} n};

[0017] Each path P i The processing time and material consumption need to be calculated. The processing time is based on the number of cutting segments and the time of each cutting segment. The time of each cutting segment is generated based on the length of each cutting segment and its corresponding feed rate. The material consumption is based on the number of cutting segments and the amount of material removed in each cutting segment. The amount of material removed in each cutting segment is generated based on the area of ​​each cutting segment and its corresponding cutting depth. The processing time and material consumption are weighted and summed to obtain the first comprehensive evaluation function. Several paths whose first comprehensive evaluation function is less than a first set threshold are selected as candidate paths.

[0018] Furthermore, the triggered secondary simulation evaluation mechanism includes the following:

[0019] Tool wear is calculated based on the number of cutting segments and the wear amount on the tool from each cutting segment. The wear amount on the tool from each cutting segment is calculated based on the wear coefficient, cutting length, and cutting depth of each cutting segment. Machining accuracy is calculated based on the number of cutting segments and the machining error of each cutting segment. The machining error of each cutting segment is calculated based on the error coefficient, cutting length, and cutting depth of each cutting segment. A second comprehensive evaluation function is obtained by weighted summation of tool wear and machining accuracy. Several candidate paths whose second comprehensive evaluation function is less than a second preset threshold are selected as qualified paths.

[0020] Furthermore, the process of executing the simulation test strategy is as follows:

[0021] Environment setup: Construct a virtual machining environment, which includes at least: machine tools, cutting tools, and fixtures;

[0022] Process simulation: A comprehensive simulation is performed on each compliance path to detect any abnormal phenomena in any cutting segment of the compliance path, and corresponding labeling operations are performed on the detection results; among them, abnormal phenomena are overcutting or collision; each compliance path includes several path segments, and each path segment consists of several cutting segments;

[0023] For cutting segments that exhibit overcutting or both overcutting and collision, the corresponding cutting segments are displayed in red.

[0024] The corresponding cutting segment where a collision occurs is displayed in yellow;

[0025] Cutting segments that do not have overcutting or collisions are displayed in green;

[0026] Results analysis: The path with the highest proportion of green markers among the compliant paths is extracted as the initial path; for the initial path, the density of non-green cutting segments in each path segment is calculated to identify high-risk segments; if no high-risk segments are identified, the initial path is used as the final path; otherwise, the path optimization mechanism is triggered.

[0027] Furthermore, the triggered path optimization mechanism is as follows: according to preset standard conditions, find other path segments from the qualified paths that can replace the current high-risk segment; after obtaining other path segments that meet the conditions, insert them into the original path to generate a new path, and recalculate its second comprehensive evaluation function; compare the second comprehensive evaluation functions of the original path and the optimized new path, and select the smaller one as the final path.

[0028] Furthermore, for the initially determined path, traverse all cutting segments S in its path segment. j Count the number of red and yellow cutting segments and calculate their proportion (D) in each path segment of the initial path. ry (P b If it is D ry (P b If the set threshold corresponding to the density of red and yellow segments is greater than the threshold, it means that the path segment is a high-risk segment and needs to be replaced and optimized; otherwise, it does not need to be replaced. The preset standard conditions are: geometric position matching: the replacement path segment should be in the same geometric area as the original path segment; functional consistency: the cutting type, direction and depth parameters are the same; low non-green cutting density: the density of non-green cutting segments in the replacement path segment is less than 1 / 2 of that in the original path segment.

[0029] This invention provides a machine vision-based simulation virtual processing system, which has the following advantages:

[0030] (1) This solution uses machine vision recognition and calibration technology to achieve precise positioning and dynamic adjustment of the target workpiece, achieving real-time feedback. It solves the problems of inefficiency and inaccuracy caused by traditional manual correction. It uses machine vision technology to capture the surface image of the target workpiece in real time, and uses the feature point matching algorithm to compare the contour features with the corresponding features in the CAD model to achieve precise positioning. Once a deviation is found, the system will immediately adjust the processing parameters to ensure that the final product meets the design requirements, which greatly improves processing efficiency and accuracy.

[0031] (2) This scheme adopts multi-level simulation and path optimization technology to achieve the selection of the optimal path, thereby reducing material waste and improving processing efficiency. It solves the problem of balancing time and cost in traditional path planning. A preliminary tool path is generated based on the shape of the target workpiece. Through the first and second simulation evaluation mechanism, the path that saves time and reduces material waste is selected. On this basis, tool wear and processing accuracy are further considered, and the path with the least wear and the highest processing accuracy is selected for processing to achieve the purpose of dual balance and effectively improve the overall processing efficiency and quality.

[0032] (3) This scheme divides the compliant path into multiple path segments and marks the cutting segments in each path segment with red, yellow and green. The system can accurately identify path segments with overcutting or collision risks and extract alternative path segments from other paths based on preset standard conditions to achieve automatic replacement of high-risk segments. This not only realizes the replacement operation at the path segment level, but also improves the robustness and adaptability of the overall path structure. The path that may have local defects becomes better than the original complete path after local replacement. The scheme proposes to perform dynamic replacement on the basis of path segments and combine it with the secondary evaluation mechanism of the comprehensive evaluation function to improve the path quality without destroying the overall structure. This not only effectively utilizes the high-quality path segments in the existing path resources, but also reduces the cost of global path replanning and improves efficiency.

[0033] (4) This solution achieves closed-loop verification of the processing path by adopting a comprehensive simulation test and path optimization technology, thus preventing potential risks. By constructing a highly realistic virtual processing environment, a comprehensive simulation test is conducted on each path to check for overcutting or collision. For problematic path segments, the system will automatically find alternative path segments to replace them and then compare them with the original path after replacement to select the optimal path. This can reduce the amount of secondary optimization operations to a certain extent, thereby improving the safety and stability of processing and enhancing work efficiency to a certain extent. Attached Figure Description

[0034] Figure 1 This is a schematic diagram illustrating the mapping of physical entities to software digital space in this invention;

[0035] Figure 2 This is a schematic diagram of the machine tool simulation state during system operation in this invention;

[0036] Figure 3 This is a schematic diagram showing the path calculation results in the present invention.

[0037] Figure 4 This is a schematic diagram of the process of building a precision virtual processing platform in this invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0039] Please see Figures 1 to 4This embodiment provides a machine vision-based simulation virtual processing system. This system combines a virtual processing flow with machine vision technology to improve the system's automation level and accuracy.

[0040] The virtual machining process is as follows: Based on DT technology, the SurfMill software has been reconstructed, connecting the software programming, production preparation, and actual machining processes. This promotes the deep integration of technology, tools, and machine tools, achieving standardized programming, transparent material usage, and standardized machine tool operation. This makes the entire production process safe and controllable, enabling precise virtual machining. In SurfMill version 9.0, the software provides a platform that can map existing material information into the software. The materials that can be selected during programming are all real materials that exist in the warehouse. When outputting the program, a corresponding process sheet is attached. The output information includes material model information, tool clamping information, and workpiece clamping information, allowing machine operators to accurately prepare and install materials based on the prompts.

[0041] SurfMill 9.0 software introduces a standardized programming process throughout the entire process, from preparation to path calculation and output. Tool holders must be added to the tools used, path programming is based on a complete machine tool model, interference checks are performed during path calculation, and collision risks are indicated in the results. Dangerous paths are prohibited from output or must be replaced. This enables path optimization and risk avoidance during the programming stage.

[0042] This virtual machining simulation system includes the following functional modules that operate sequentially:

[0043] I. Model Import and Preprocessing Module:

[0044] Before starting any processing task, it is necessary to accurately import the three-dimensional model of the target workpiece into this system. This step not only requires high accuracy, but also requires consideration of how to effectively manage and organize these model data.

[0045] According to the existing SurfMill 9.0 user guide, when importing 3D models, users can enter the template type selection interface through the "New" command, and then select the appropriate environment (such as surface machining) to import the workpiece, blank, and fixture models. During this process, it is important to ensure that all models are correctly classified and assigned to their respective layers so that subsequent operations are more convenient and efficient.

[0046] The specific implementation details are as follows:

[0047] The imported 3D model is converted to a new format, potential problem sources are detected, the repair solutions provided in the solution library are compared, the repair process is completed, and the 3D model is standardized.

[0048] The system supports various 3D file formats, including at least STL and STEP. For example, when importing models from external CAD software, users can select the appropriate file format. To ensure compatibility, the system includes a built-in automatic conversion tool to convert uncommon file formats to a standard format recognizable by the system. For instance, it can convert OBJ files to STL format for easier subsequent processing. The system also features automatic detection and repair functions. For imported 3D models, the system automatically detects potential problems, i.e., potential problem sources, including non-manifold edges, self-intersecting surfaces, and topological errors. The system provides corresponding repair solutions from a benchmark library, including non-manifold edge repair solutions, self-intersecting surface repair solutions, and geometric topology repair solutions. Finally, the system standardizes the units used in imported 3D models to ensure data consistency. For example, if a workpiece model uses millimeters while a fixture model uses inches, the system will automatically convert the units during import to ensure both are operated in the same coordinate system.

[0049] Specifically, the detailed descriptions of the repair solutions provided in the benchmarking solution library are as follows:

[0050] The non-manifold edge repair scheme works as follows: Non-manifold edges refer to edges that do not meet manifold conditions, i.e., an edge connects more than two faces or does not belong to any face. In this case, the model cannot correctly represent the solid object, affecting the accuracy of subsequent processing. Therefore, the corresponding repair process is: the system traverses all edges, checking whether each edge connects more than two faces or does not belong to any face; for detected non-manifold edges, the repair action is achieved by merging adjacent faces or deleting redundant edges; in addition, it also includes closing holes in the 3D model by adding new faces. The self-intersecting surface repair scheme works as follows: Self-intersecting surfaces refer to surfaces in the model that intersect with themselves or other surfaces. In this case, errors will occur in rendering and subsequent processing; detecting self-intersecting surfaces usually involves geometric calculations, i.e., Boolean operations or ray casting; therefore, the corresponding repair process is: using ray casting or Boolean operations to check for self-intersecting surfaces in the 3D model. Does a self-intersecting surface exist? Upon discovery of a self-intersecting surface, different repair strategies are adopted based on the specific situation. The repair actions used in this embodiment include cutting and recombining the self-intersecting parts, or directly deleting the problematic parts and filling in the blank areas. The geometric topology repair scheme's technical principle is as follows: a method for checking and verifying the structural integrity of a 3D model, ensuring the validity of the 3D model by analyzing the relationships between vertices, edges, and faces in the model. For example, a valid 3D model does not contain isolated vertices or dangling edges. Therefore, the corresponding repair process is: traversing the 3D model and checking whether each vertex, edge, and face conforms to the expected topological rules. For example, each edge should connect exactly two faces, and each vertex should connect at least three edges. If a violation of the rules is found, the corresponding repair action is triggered: for isolated vertices, they are deleted; for dangling edges, these edges are covered by expanding adjacent faces.

[0051] By adopting an automatic detection and repair technology solution, high-precision 3D model import is achieved, which not only achieves data standardization but also solves the processing deviation problem caused by model errors in traditional technologies. In traditional processing workflows, problems such as non-manifold edges and self-intersecting surfaces that may exist during 3D model import can lead to decreased accuracy or even failure in subsequent processing. This solution introduces automatic detection and repair functions to identify and repair these problems during the model import stage, ensuring the integrity and accuracy of the model. At the same time, by automatically converting files of different formats and unifying units, data consistency is further improved, thus laying a solid foundation for subsequent processing path planning.

[0052] II. Machine Vision Recognition and Calibration Module:

[0053] Machine vision technology is used to locate and capture the target workpiece and match its features. When a deviation is detected, an adjustment and re-calibration mechanism is triggered to ensure that the target workpiece meets the design requirements.

[0054] The specific details of the operating scheme in the machine vision recognition and calibration module are as follows:

[0055] Positioning and capture: A network of vision sensors mounted on the machine tool is used to capture real-time images of the target workpiece's surface, and edge detection algorithms are employed to identify contour features; the machine tool represents the equipment used to process the target workpiece, and is referenced... Figure 2 This can be seen from the following: Visual sensor network: Several high-resolution cameras installed around the target workpiece on the machine tool form a comprehensive visual monitoring network, used to capture every detail of the target workpiece from all angles without blind spots; Feature matching: A feature point matching algorithm is used to compare the contour features with the corresponding features in the 3D model, thereby achieving precise positioning; Adjustment and re-verification mechanism: Once a deviation is detected, the preset machining parameters are immediately adjusted (the principle is the same as the benchmarking scheme library mentioned in the model import and preprocessing module, so it will not be elaborated here), and a calibration notification is issued, prompting a secondary calibration action to complete the re-verification or secondary adjustment, so as to ensure that the final product meets the design requirements; For example, if the target workpiece is detected to be offset by 0.5 mm, the corresponding scheme in the scheme library is immediately retrieved, that is, the tool path is automatically adjusted to ensure that the machining accuracy is not affected; In a practical case, assuming that a slight offset of the target workpiece is detected when machining a precision workpiece mold, the tool path is immediately adjusted and the operator is notified to perform a secondary calibration to ensure the quality of the final product.

[0056] It should be noted that the machine vision recognition and calibration module greatly enhances the system's adaptability, enabling it to maintain high processing accuracy even in the face of complex and ever-changing working conditions. Since the specific adjustments and re-verifications are not the focus of this embodiment, and are existing solutions, the adjustment and re-verification calibration mechanisms will not be elaborated upon.

[0057] By adopting machine vision recognition and calibration technology, precise positioning and dynamic adjustment of the workpiece are achieved. This not only provides real-time feedback but also solves the problems of inefficiency and inaccuracy associated with traditional manual calibration. In traditional machining processes, the actual position of the workpiece after clamping often requires manual measurement and calibration, which is not only time-consuming and labor-intensive but also prone to human error. This solution utilizes machine vision technology to capture real-time images of the workpiece surface through a high-resolution camera network installed on the machine tool. It then uses a feature point matching algorithm to compare the contour features with the corresponding features in the CAD model to achieve precise positioning. Once a deviation is detected, the system immediately adjusts the machining parameters to ensure that the final product meets the design requirements, greatly improving machining efficiency and accuracy.

[0058] III. Processing Path Planning Module:

[0059] Based on the accurate information provided by the first two modules, this module focuses on generating the optimal toolpath;

[0060] The specific implementation details are as follows:

[0061] Based on the shape of the target workpiece and the processing target (i.e. the required processing area), several feasible toolpaths are initially set. A simulation evaluation mechanism is executed on different toolpath schemes to select candidate paths from the several feasible toolpaths, and a second simulation evaluation mechanism is triggered to extract the qualified path from the candidate paths.

[0062] The initial path setting is based on the target workpiece geometry and machining objectives. The system automatically generates several feasible toolpaths by identifying boundary contours, feature areas (such as bosses and grooves), and machining depths. In this embodiment, commonly used strategies in actual machining include parallel cutting, circumferential cutting, or reciprocating paths to ensure coverage of the entire machining area. In practical applications, for some relatively conventional rectangular workpieces, the system generates equidistant straight paths along the X / Y directions based on the length, width, and machining allowance, and adds transition arcs at corners to ensure continuity and machining stability. Each path consists of several cutting segments, and the system automatically marks key points (such as the starting point, reversal point, and tool entry position) to facilitate subsequent optimization and simulation analysis.

[0063] The simulation evaluation mechanism implemented is as follows:

[0064] Given a set of paths P = {P1, P2, ..., P...} n}; i = 1, 2, ..., n, where n is a positive integer and i is the corresponding path number;

[0065] Among them, each path P i It is necessary to calculate its processing time and material consumption; processing time T(P) i The formula for calculating processing time is: ; where m i It is path P i Number of cutting stages, t j The time of the j-th cutting segment is based on: t j =l j / v j ;l j v is the length of the j-th cutting segment. j This refers to the feed rate of the cutting section; material consumption M(P) i The formula for calculating material consumption is: ; where mlj It is the material removal amount of the j-th cutting segment, based on: ml j =A j / d j Among them, A j d is the area of ​​the j-th segment being cut. j This is the cutting depth of this segment; First comprehensive evaluation function: First comprehensive evaluation function F1(P) i The result is derived from a weighted sum of processing time and material consumption.

[0066] Where w1 and w2 are the weighting coefficients for processing time and material consumption, respectively, with values ​​ranging from 0 to 1; Filtering path: Select the path that makes F1(P i Several paths whose values ​​are less than a first set threshold are selected as candidate paths; where the first set threshold is set according to actual needs, and is usually selected by filtering out paths that make F1(P) less than a certain threshold. i The path that is the smallest or closest to the smallest.

[0067] It should be noted that the result of executing the above simulation evaluation mechanism once is as follows:

[0068] Based on selecting several paths that save time and reduce material waste, we need to further consider tool wear and machining accuracy on different paths to ensure a balance between the two. Therefore, we need to implement a subsequent mechanism.

[0069] The triggered secondary simulation evaluation mechanism includes the following:

[0070] Tool wear W(P) i The formula for calculating tool wear is: Among them, wl j The wear on the tool from the j-th cutting segment is estimated using the formula: wl j =k j ×l j ×d j ; where k j Let l represent the wear coefficient of the j-th cutting segment. j d represents the cutting length of the j-th cutting segment. j This indicates the depth of cut for that section; machining accuracy Q(P) i Formula for calculating machining accuracy: ;q j It is the machining error of the j-th cutting segment; estimated by this formula: q j =e j ×l j ×d j , where e j Let l represent the error coefficient of the j-th cutting segment. jThis represents the cutting length of the segment, d. j This represents the depth of cut in the cutting segment; Second comprehensive evaluation function: Second comprehensive evaluation function F2(P) i The result is derived by weighted summation based on tool wear and machining accuracy.

[0071] ; where w3 and w4 are the weighting coefficients for tool wear and machining accuracy, respectively, with values ​​ranging from 0 to 1; two-screen path: select several candidate paths that make F2(Pi) less than the second set threshold as the standard path; where the second set threshold and the first set threshold are also set according to actual needs, and usually the path that makes F2(Pi) the smallest or close to the smallest is selected.

[0072] By employing a multi-level simulation and path optimization technology, the optimal path selection is achieved, thus reducing material waste and improving processing efficiency. This also solves the problem of balancing time and cost in traditional path planning. Traditional path planning methods typically rely on experienced technicians manually setting paths or based on historically established paths, making it difficult to balance processing efficiency and material consumption. This solution first generates a preliminary toolpath based on the target workpiece shape. Then, through primary and secondary simulation evaluation mechanisms, it selects the path that saves both time and reduces material waste. Furthermore, considering tool wear and processing accuracy, it selects the path with the least wear and highest processing accuracy, thereby achieving a dual balance and effectively improving overall processing efficiency and quality.

[0073] IV. Virtual Simulation and Optimization Module:

[0074] After completing path planning, the system will enter the simulation phase, execute the simulation test strategy, test the processing target, and decide whether to trigger the path optimization mechanism based on the test results to generate the final path;

[0075] The simulation test strategy implemented is as follows:

[0076] S101. Environment Setup: Construct a highly realistic virtual machining environment, requiring precise modeling of all components, including machine tools, cutting tools, and fixtures; among which, referencing Figure 4 It can be seen that precise explanations are given regarding material standardization, material digitization, and programming standardization;

[0077] S102. Process simulation: Perform an overall simulation for each compliance path, detect any abnormal phenomena in any cutting segment of the compliance path, and perform corresponding marking operations on the detection results.

[0078] Among them, the abnormal phenomena are overcutting or collision; each qualified path includes several path segments, and each path segment consists of several cutting segments; for example, if a qualified path is shaped like a square, then there are four path segments, and each path segment can be set to five cutting segments; since a path segment replacement operation needs to be introduced later, the original qualified path needs to be replaced.

[0079] For details, please refer to the appendix. Figure 3 As shown, the system interface relates to the path groups formed by the compliant paths:

[0080] Cutting segments with overcutting or collision are displayed in red; cutting segments with collision are displayed in yellow; and cutting segments without overcutting or collision are displayed in green. In this embodiment, after extensive simulation, it was found that there is a significant correlation between red and yellow density and subsequent tool wear and machining accuracy. Therefore, red and yellow density is not only a visual anomaly indicator, but also an important indicator for predicting machining performance.

[0081] S103. Extract the path with the largest proportion of green markers from the qualified paths as the initial path;

[0082] S104. For the initial path, calculate the density of non-green cutting segments in each path segment to identify high-risk segments; if no high-risk segments are identified, the initial path is used as the final path.

[0083] If a high-risk segment is identified, the path optimization mechanism is triggered: return to the previous level, and find other path segments from the qualified paths that can replace the current high-risk segment according to preset standard conditions; after obtaining other path segments that meet the conditions, insert them into the original path to generate a new path, and recalculate its second comprehensive evaluation function; compare the second comprehensive evaluation function of the original path with that of the optimized new path, and select the smaller one as the final path;

[0084] For the initially determined path, all cutting segments S in its path segment are traversed. j Count the number of red and yellow cutting segments and calculate their proportion in each path segment of the initial path: ; where N red (P b ), N yellow (P b ) and N total (P b These represent: the number of cutting segments marked in red, the number of cutting segments marked in yellow, and the total number of cutting segments in any segment of the initially defined path; if D ry (P b > The set threshold T corresponding to the density of the red and yellow segments ryIf the value is positive, it indicates that the path segment is a high-risk segment and needs to be replaced or optimized; otherwise, it does not.

[0085] Preset standard conditions are expressed as follows:

[0086] Geometric location matching: The alternative path segment should be located in the same or similar geometric area as the original path segment;

[0087] Consistent in function: The cutting type (roughing / finishing), direction, and depth parameters are the same or similar;

[0088] Low density of non-green cutting segments: The density of non-green cutting segments in the alternative path segment should be less than 1 / 2 of that in the original path segment;

[0089] Among them, similar parameters mean that the corresponding parameters are within the error range;

[0090] Alternative path segments are screened based on three dimensions: geometric location matching, functional consistency, and low non-green density, ensuring that the replaced paths meet the requirements in terms of function, structure, and security. During actual operation, the system discovered that some paths excluded due to poor overall performance actually have segments with extremely high substitutable value. Therefore, this embodiment uses the aforementioned preset standard conditions for screening, no longer pursuing perfection for every path, but focusing on the reuse potential of path segments. This solution breaks with this conventional thinking, proposing the concept of "path segment availability priority," reconstructing the path resource management method from a micro perspective, and achieving a breakthrough and optimization of the current technological barriers.

[0091] By comparing the second comprehensive evaluation function of the original path P0 (i.e. the initial path) and the optimized path P0', the path with the smaller value is selected as the final path. Through the above process, high-risk path segments can be accurately identified and efficiently replaced, thereby improving the safety and stability of the overall path while ensuring processing quality.

[0092] By dividing the compliant path into multiple path segments and marking the cutting segments within each segment with red, yellow, and green labels, the system can accurately identify path segments with overcutting or collision risks. Based on preset standard conditions, it extracts alternative path segments from other paths, achieving automatic replacement of high-risk segments. This not only realizes path segment-level replacement operations but also improves the robustness and adaptability of the overall path structure. As a result, paths that might have local defects become better than the original complete paths after local replacement. It should be noted that traditional path optimization often adopts a global replacement strategy, assuming that the path is an indivisible whole, and replacing the entire path is the current mainstream approach. However, the solution presented in this embodiment proposes for the first time to perform dynamic replacement at the path segment level and combine it with a secondary evaluation mechanism based on a comprehensive evaluation function. This improves path quality without destroying the overall structure, effectively utilizing high-quality path segments in existing path resources, reducing the cost of global path replanning, and improving efficiency.

[0093] The advantages of implementing the path optimization mechanism in the simulation testing strategy are as follows: 1. Reasonable path segment division: ensuring that each path segment is independent and replaceable; 2. Strict selection of alternative segments: geometric position, cutting function, and red / yellow density must all be matched; 3. Comprehensive performance evaluation: comprehensively considering tool wear and machining accuracy; 4. High degree of automation: the entire process can be automatically executed by the system, and automatically complete the action or operation of confirming the alternative strategy; In addition, machine learning algorithms can be introduced on the basis of this mechanism to train the model based on historical data, automatically predict which path segments are more likely to have red / yellow problems, thereby avoiding these areas in the path planning stage, realizing proactive risk avoidance path optimization, and further improving the level of intelligent manufacturing.

[0094] The final path is then further optimized to generate the optimal path.

[0095] The secondary optimization process includes: reducing the feed rate, changing the cutting direction, increasing coolant injection, and using an adaptive control algorithm; under the conditions of the simulation phase:

[0096] Reduce feed rate: In cutting segments where overcutting or collision occurs, appropriately reduce the feed rate to minimize the impact of the tool on the workpiece; Example: Suppose an overcutting problem is detected in a certain cutting segment, the system reduces the feed rate of that segment from 500 mm / min to 300 mm / min. mm / min, after resimulation, the overcutting problem disappeared; Changing the cutting direction: Changing the cutting direction of the tool to avoid collision with the target workpiece; Example: Suppose that there is a collision problem in a certain path segment during the machining of a complex curved surface, the system changes the cutting direction of this path from clockwise to counterclockwise, and after resimulation, the collision problem disappears; Adding coolant spray: In high-temperature areas or areas prone to overcutting, add coolant spray function to reduce temperature and reduce overcutting; Example: Suppose that there is an overcutting problem in a certain path segment when machining titanium alloy blades, the system adds coolant spray function to this path segment, and after resimulation, the overcutting problem disappears; Adaptive control algorithm: Using an adaptive control algorithm, the tool's motion trajectory is dynamically adjusted according to the changes in the target workpiece surface to ensure machining quality and efficiency; Example: Suppose that there is an overcutting problem in a certain path segment when machining a complex curved surface, the system uses an adaptive control algorithm to dynamically adjust the tool's motion trajectory according to the changes in the curved surface, and after resimulation, the overcutting problem disappears.

[0097] By adopting a comprehensive simulation testing and path optimization technology solution, closed-loop verification of the processing path was achieved, which not only prevented potential risks, but also solved the problems of frequent rework and resource waste in the traditional trial cutting process.

[0098] In traditional processing, the lack of effective simulation verification methods often necessitates multiple trial cuts to find a suitable processing path, leading to resource waste and extended production cycles. This solution constructs a highly realistic virtual processing environment to conduct comprehensive simulation tests on each path, checking for overcutting or collisions. For problematic path segments, the system automatically finds alternative path segments for replacement and performs a secondary comparison with the original path after replacement, thereby selecting the optimal path. This reduces the amount of secondary optimization operations to a certain extent, improving processing safety and stability while also enhancing work efficiency.

[0099] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0100] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A machine vision-based simulation virtual processing system, characterized in that, include: The model import and preprocessing module imports the 3D model of the target workpiece, performs format conversion, detects potential problem sources, compares it with the repair solutions given in the solution library, completes the repair process, and completes data standardization. The machine vision recognition and calibration module uses machine vision technology to locate and capture the target workpiece and match its features. When a deviation is detected, it triggers an adjustment and recalibration mechanism to ensure that the target workpiece meets the design requirements. The machining path planning module initially sets several toolpaths based on the shape of the target workpiece and the machining objectives. It then performs a simulation evaluation mechanism on different toolpath schemes to filter candidate paths and triggers a secondary simulation evaluation mechanism. This mechanism involves: calculating tool wear based on the number of cutting segments and the wear amount on the tool from each segment (the wear amount is calculated based on the wear coefficient, cutting length, and cutting depth of each segment); calculating machining accuracy based on the number of cutting segments and the machining error of each segment (the machining error is calculated based on the error coefficient, cutting length, and cutting depth of each segment); and finally, deriving a second comprehensive evaluation function by weighted summation of tool wear and machining accuracy. Candidate paths whose second comprehensive evaluation function is less than a second preset threshold are then selected as qualifying paths. The virtual simulation and optimization module, after completing path planning, enters the simulation phase, executes a simulation testing strategy, tests the processing target, and determines whether to trigger the path optimization mechanism based on the test results to generate the final path. It then performs secondary optimization on the final path to generate the optimal path. The triggering of the path optimization mechanism involves: according to preset standard conditions, finding alternative path segments from the qualified paths that can replace the current high-risk segment; after obtaining other suitable path segments, inserting them into the original path to generate a new path, and recalculating its second comprehensive evaluation function; comparing the second comprehensive evaluation functions of the original path and the optimized new path, and selecting the smaller one as the final path.

2. The machine vision-based simulation virtual processing system according to claim 1, characterized in that: When performing format conversion, the supported file formats include at least STL and STEP; when detecting potential problem sources, the potential problem sources include at least non-manifold edges, self-intersecting surfaces, and topological errors; the corresponding repair solutions provided by the benchmark solution library include at least non-manifold edge repair solutions, self-intersecting surface repair solutions, and geometric topology repair solutions; data standardization processing: the imported 3D model is standardized in terms of size units.

3. The machine vision-based simulation virtual processing system according to claim 1, characterized in that: The process of locating and capturing the target workpiece is as follows: by using a pre-installed network of vision sensors to capture the surface image of the target workpiece in real time, and using an edge detection algorithm to identify contour features; The feature matching process is as follows: a feature point matching algorithm is used to compare the contour features with the corresponding features in the 3D model to complete the positioning operation.

4. The machine vision-based simulation virtual processing system according to claim 3, characterized in that: The triggered adjustment and recalibration mechanism process is as follows: when a deviation is detected, the preset processing parameters are adjusted and a calibration notification is issued to prompt a second calibration action to complete the recalibration or second adjustment.

5. The machine vision-based simulation virtual processing system according to claim 1, characterized in that: The initial path setting is based on the target workpiece geometry and machining target. By identifying the boundary contour, feature area and machining depth, several tool paths are generated. The machining strategies employed include at least: parallel cutting, circumferential cutting, and reciprocating paths.

6. The machine vision-based simulation virtual processing system according to claim 1, characterized in that: The process of executing a simulation evaluation mechanism is as follows: Suppose there is a set of paths P = {P1, P2, ..., Pn}; Each path Pi needs to have its processing time and material consumption calculated. The processing time is determined based on the number of cutting segments and the time of each cutting segment, while the time of each cutting segment is generated based on the length of each cutting segment and its corresponding feed rate. The material consumption is determined based on the number of cutting segments and the amount of material removed by each cutting segment, while the amount of material removed by each cutting segment is generated based on the area of ​​each cutting segment and its corresponding cutting depth. The processing time and material consumption are weighted and summed to obtain a first comprehensive evaluation function. Several paths whose first comprehensive evaluation function is less than a first set threshold are selected as candidate paths.

7. The machine vision-based simulation virtual processing system according to claim 1, characterized in that: The execution of the simulation test strategy is as follows: Environment setup: Construct a virtual machining environment, which includes at least: machine tools, cutting tools, and fixtures; Process simulation: A comprehensive simulation is performed on each compliance path to detect any abnormal phenomena in any cutting segment of the compliance path, and corresponding labeling operations are performed on the detection results; among them, abnormal phenomena are overcutting or collision; each compliance path includes several path segments, and each path segment consists of several cutting segments; For cutting segments that exhibit overcutting or both overcutting and collision, the corresponding cutting segments are displayed in red. The corresponding cutting segment where a collision occurs is displayed in yellow; Cutting segments that do not have overcutting or collisions are displayed in green; Results analysis: The path with the highest proportion of green markers among the compliant paths is extracted as the initial path; for the initial path, the density of non-green cutting segments in each path segment is calculated to identify high-risk segments; if no high-risk segments are identified, the initial path is used as the final path; otherwise, the path optimization mechanism is triggered.

8. The machine vision-based simulation virtual processing system according to claim 7, characterized in that: For the initial path, traverse all cutting segments Sj in its path segment, count the number of red and yellow cutting segments, and calculate their proportion Dry(Pb) in each path segment of the initial path. If Dry(Pb) > the set threshold corresponding to the density of red and yellow segments, it means that the path segment is a high-risk segment and needs to be replaced and optimized; otherwise, it does not need to be replaced. The preset standard conditions are: geometric position matching: the replacement path segment should be in the same geometric area as the original path segment; functional consistency: the cutting type, direction, and depth parameters are the same. Low non-green cutting density: The density of non-green cutting segments in the alternative path segment is less than 1 / 2 of that in the original path segment.

Images