Simulation virtual processing system based on machine vision

Abstract

The invention discloses a simulation virtual processing system based on machine vision, and relates to the technical field of machine vision. The machining system comprises a model importing and preprocessing module, a machine vision recognition and calibration module, a machining path planning module and a virtual simulation and optimization module. The technical key points are as follows: by adopting a comprehensive simulation test and path optimization technical scheme, closed-loop verification of a processing path is realized, an effect of preventing potential risks is achieved, and by constructing a highly vivid virtual processing environment, comprehensive simulation test is performed on each path, and whether an overcut or collision phenomenon exists or not is checked; for path sections with problems, the system searches for replaceable path sections for replacement, and secondary comparison with the original path is carried out after replacement, so that the optimal path is screened out, the operation content of secondary optimization processing can be reduced to a certain extent, on one hand, the processing safety and stability are improved, and on the other hand, the processing efficiency is improved. And on the other hand, the working efficiency is improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of machine vision, and in particular to a simulation virtual machining system based on machine vision. Background Art

[0002] Machine vision aims to enable computers or machines to see and understand their surroundings like humans. Specifically, machine vision systems use cameras or other image acquisition devices to capture images of target objects, 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, identification and tracking, guiding robots, and virtual processing simulation.

[0003] Traditional virtual machining simulations present several significant challenges in CNC machining path planning and optimization. First, for precise workpiece positioning and dynamic adjustment, traditional methods rely on manual calibration, which is inefficient and prone to human error. For example, in the machining of complex curved parts, technicians spend a significant amount of time measuring and calibrating, but even then, high-precision positioning is difficult to achieve, resulting in the final product not meeting design requirements. Second, during the path planning stage, traditional technologies often struggle to balance machining time and material consumption. Path planning typically relies on pre-programmed, established paths. Even if this approach can balance machining efficiency, it fails to effectively balance other aspects, such as machining accuracy. Third, traditional path planning methods typically employ a global replacement strategy, assuming that the path as a whole cannot be disassembled. Replacing the entire path is the prevailing approach. This results in local defects potentially causing the entire path to be discarded, increasing the cost and time of replanning. For example, when machining large mold inserts, if a problem exists in a particular path segment, the entire path must be redesigned or replaced with another path that requires verification. This process can result in significant resource waste and extend production cycles. Summary of the Invention

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: A simulated virtual machining system based on machine vision, comprising: The model import and preprocessing module imports the 3D model of the target workpiece, performs format conversion processing, detects potential problem sources, compares the repair solutions given in the solution library, completes the repair process, and completes data standardization processing; The machine vision recognition and calibration module uses machine vision technology to locate, capture, and match features of the target workpiece. When deviations are identified, an adjustment and recalibration mechanism is triggered to ensure that the target workpiece meets the design requirements. The machining path planning module preliminarily sets several tool paths based on the shape of the target workpiece and the machining objectives. It then performs a primary simulation evaluation mechanism on different tool path schemes to screen out candidate paths and triggers a secondary simulation evaluation mechanism to extract the qualified path from the candidate paths. 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.

[0005] Furthermore, when performing format conversion processing, multiple file formats are supported, including 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 repair solutions given by the benchmark solution library correspond to at least: non-manifold edge repair solutions, self-intersecting surface repair solutions and geometric topology repair solutions; data standardization processing: the imported three-dimensional model is unified in size units.

[0006] Furthermore, the process of locating and capturing the target workpiece is as follows: by using a pre-installed visual 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 three-dimensional model to complete the positioning operation.

[0007] Furthermore, the triggered adjustment and re-inspection calibration mechanism process is: when a deviation is identified, the preset processing parameters are adjusted, and a calibration notification is issued to prompt a secondary calibration action to complete the re-inspection or secondary adjustment.

[0008] Furthermore, the preliminary path setting is based on the target workpiece geometry and machining objectives by identifying the boundary contour, feature area and machining depth to generate several tool paths; Among them, the strategies adopted during processing include at least: parallel cutting, circumferential cutting and reciprocating path.

[0009] Furthermore, the process of executing a simulation evaluation mechanism is as follows: Suppose there is a set of paths P={P1, P2, ..., P n}; Each path P iIt is necessary to calculate its processing time and material consumption; wherein, the processing time is obtained based on the number of cutting segments and the time of each cutting segment, and the time of each cutting segment is generated based on the length of each cutting segment and its corresponding feed speed; the material consumption is obtained based on the number of cutting segments and the material removal amount of each cutting segment, and the material removal amount of each cutting segment is generated based on the area of each cutting segment and its corresponding cutting depth; the first comprehensive evaluation function is obtained by weighted summation of the processing time and material consumption; several paths whose first comprehensive evaluation function is less than the first set threshold are screened out and used as candidate paths.

[0010] Furthermore, the triggered secondary simulation evaluation mechanism is as follows: The tool wear is calculated based on the number of cutting segments and the amount of tool wear caused by each cutting segment. The amount of tool wear caused by each cutting segment is calculated based on the wear coefficient, cutting length and cutting depth of each cutting segment. The 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 generated 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 taken as qualified paths.

[0011] Furthermore, the simulation test strategy process is as follows: Environment construction: Build a virtual machining environment, including at least: machine tools, cutting tools and fixtures; Process simulation: A comprehensive simulation is performed on each path to detect any anomalies in any cutting segment along the path, and corresponding identification is performed on the detection results. Anomalies include overcutting or collisions. Each path consists of several path segments, and each path segment is composed of several cutting segments. The corresponding cutting segments with overcutting, overcutting and collision are displayed in red; The corresponding cutting segments where there is a collision are displayed in yellow; The corresponding cutting segments without overcutting or collision are displayed in green; Result analysis: The path with the largest proportion of green markings among the qualified paths is extracted as the preliminary path. For the preliminary 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 preliminary path is used as the final path. Otherwise, the path optimization mechanism is triggered.

[0012] Furthermore, the triggered path optimization mechanism includes the following: according to preset standard conditions, other path segments that can replace the current high-risk segment are found from the qualified path; after obtaining other path segments that meet the conditions, they are inserted into the original path to generate a new path, and its second comprehensive evaluation function is recalculated; the second comprehensive evaluation function of the original path and the optimized new path are compared, and the smaller one is selected as the final path.

[0013] Furthermore, for the initial 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 D in each path segment in the initial path ry (P b ), if D ry (P b )>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 indicate: 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 the original path segment.

[0014] The present invention provides a machine vision-based simulation virtual machining system, which has the following beneficial effects: (1) This solution uses machine vision recognition and calibration technology to achieve precise positioning and dynamic adjustment of the target workpiece, achieving the effect of real-time feedback and solving the inefficiency and inaccuracy problems caused by traditional manual correction. It uses machine vision technology to capture the surface image of the target workpiece in real time and uses 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, greatly improving the processing efficiency and accuracy. (2) This solution uses multi-level simulation and path optimization technology to achieve the selection of the optimal path, achieving the effect of reducing material waste and improving processing efficiency, and solving the problem of difficult to balance time and cost in traditional path planning. A preliminary tool path is generated according to the target workpiece shape, and a path that saves time and reduces material waste is screened through the primary and secondary simulation evaluation mechanism. 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 double balance and effectively improve the overall processing efficiency and quality. (3) This solution divides the qualified path into multiple path segments and marks the cutting segments in red, yellow and green in each path segment. The system can accurately identify the 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. It not only realizes the replacement operation at the path segment level, but also improves the robustness and adaptability of the overall path structure, so that the path that may have local defects is better than the original complete path after local replacement. It proposes dynamic replacement based on path segments and combines the secondary evaluation mechanism of comprehensive evaluation function to improve the path quality without destroying the overall structure. It 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. (4) This solution achieves closed-loop verification of the machining path by adopting comprehensive simulation testing and path optimization technology, thus preventing potential risks. By constructing a highly realistic virtual machining environment, a comprehensive simulation test is performed on each path to check whether there is overcutting or collision. For problematic path segments, the system will automatically search for alternative path segments to replace them, and perform a secondary comparison with the original path after replacement to select the optimal path. This can reduce the operation content of the secondary optimization process to a certain extent, which can improve the safety and stability of the machining on the one hand, and enhance work efficiency to a certain extent on the other. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic diagram of mapping a physical entity to a software digital space in the present invention; Figure 2 This is a schematic diagram of the machine tool simulation state when the system is running in the present invention; Figure 3 This is a schematic diagram of the path calculation result display state in the present invention; Figure 4 Schematic diagram of the process of building a precise virtual processing platform in the present invention. DETAILED DESCRIPTION

[0016] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] See also Figures 1 to 4 ,This embodiment provides a simulated virtual machining system based on machine vision, which will combine the virtual machining process and introduce machine vision technology to improve the automation level and accuracy of the system; The virtual machining process is as follows: SurfMill software has been reconstructed based on DT technology, connecting the processes of software programming, production preparation and actual machining, promoting the deep integration of technology, tools and machine tools, and realizing programming standardization, transparent material use and standardized machine tool operation, so as to make the entire production process safe and controllable and realize precise virtual machining; in the SurfMill9.0 version, the software provides a platform that can map the actual existing material information to the software. The materials that can be selected during programming are all real materials in the warehouse, and a corresponding process sheet is attached when the program is output. The output information includes material model information, tool clamping information and workpiece clamping information, etc. The operator can accurately prepare and install materials according to the prompt information; SurfMill 9.0 introduces a standardized programming process from the preparation stage to path calculation and output. Tools must be equipped with toolholders, and path programming is based on a complete machine model. Interference checks are performed during path calculation, and collision risks are indicated in the results. Dangerous paths cannot be output or changed. This allows for path optimization and risk avoidance during the programming phase. The simulation virtual machining system includes the following functional modules that run in sequence: 1. Model import and preprocessing module: Before starting any machining task, the 3D model of the target workpiece needs to be accurately imported into the system. This step requires not only high accuracy but also effective management and organization of the model data. According to the existing SurfMill 9.0 operating guide, when importing a 3D model, users can use the "New" command to enter the template type selection interface, and then select the appropriate environment (such as surface processing) 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 to make subsequent operations more convenient and efficient. The specific implementation contents are: Convert the imported 3D model to a new format, detect potential problem sources, compare the repair solutions provided by the solution library, complete the repair process, and perform data standardization on the 3D model. Among them, file format conversion: this system supports a variety of mainstream 3D file formats, including at least: STL and STEP; for example, when importing models from external CAD software, users can choose the corresponding file format for import; in order to ensure compatibility, the system can have built-in automatic conversion tools to convert uncommon file formats into standard formats that the system can recognize; for example, converting OBJ files to STL format to facilitate subsequent processing; automatic detection and repair function: for imported three-dimensional models, the system can automatically detect potential problems, that is, potential problem sources, including non-manifold edges, self-intersecting surfaces and topological errors; the repair solutions given by the benchmarking solution library correspond to non-manifold edge repair solutions, self-intersecting surface repair solutions and geometric topology repair solutions; data standardization processing: the imported three-dimensional models are unified in size units to ensure data consistency; for example, in actual operation, assuming that a workpiece model uses millimeters and the other fixture model uses inches, the system will automatically convert units during import to ensure that the two are operated in the same coordinate system.

[0018] Specifically, the detailed introduction of the given repair solutions in the benchmark solution library is as follows: The technical principle of the non-manifold edge repair solution is: non-manifold edges refer to those edges that do not meet the manifold conditions, that is, 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 subsequent processing accuracy; therefore, the corresponding repair process is: the system traverses all edges and checks whether each edge connects more than two faces or does not belong to any face; for the detected non-manifold edges, the repair processing action is achieved by merging adjacent faces or deleting redundant edges; in addition, it also includes closing holes in the three-dimensional model by adding new faces; the technical principle of the self-intersecting face repair solution is: self-intersecting faces refer to certain faces in the model that intersect with themselves or other faces. In this case, errors will occur in the rendering and subsequent processing process; detecting self-intersecting faces usually involves geometric calculations, that is, Boolean operations or ray casting methods; therefore, the corresponding repair process is: use ray casting methods or Boolean operations to check whether there are self-intersecting faces in the three-dimensional model. Whether there are self-intersecting surfaces; after discovering self-intersecting surfaces, different repair strategies are adopted according to the specific situation; the repair processing actions adopted in this embodiment include cutting and recombining self-intersecting parts, or directly deleting problematic parts and filling blank areas; the geometric topology repair solution, its technical principle is: a method for checking and verifying the structural integrity of the three-dimensional model, by analyzing the relationship between the vertices, edges, and faces in the model to ensure the validity of the three-dimensional model; for example, a valid three-dimensional model does not contain isolated vertices or dangling edges; therefore, the corresponding repair process is: traverse the three-dimensional model to check 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 processing action is triggered: for isolated vertices, delete them; for dangling edges, cover these edges by extending adjacent faces.

[0019] By adopting the automatic detection and repair technology solution, high-precision 3D model import is achieved, which not only achieves the effect of data standardization, but also solves the processing deviation problem caused by model errors in traditional technologies. In the traditional processing process, problems such as non-manifold edges and self-intersecting surfaces that may exist when importing 3D models will lead to a decrease in accuracy or even failure in the subsequent processing process. By introducing automatic detection and repair functions, this solution identifies and repairs 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, the consistency of the data is further improved, thus laying a solid foundation for subsequent processing path planning.

[0020] 2. Machine vision recognition and calibration module: Machine vision technology is used to locate, capture, and match features of the target workpiece. When deviations are identified, an adjustment and recalibration mechanism is triggered to ensure that the target workpiece meets the design requirements. The specific description of the operation scheme in the machine vision recognition and calibration module is as follows: Positioning capture: By using the visual sensor network installed on the machine tool to capture the surface image of the target workpiece in real time, the edge detection algorithm is used to identify the contour features; the machine tool represents the equipment used to process the target workpiece, and the machine tool is used to identify the contour features by referring to the machine tool. Figure 2 As can be seen, the visual sensor network consists of several high-resolution cameras installed on the machine tool around the target workpiece, forming a comprehensive visual monitoring network for capturing every detail of the target workpiece in all directions without blind spots. Feature matching involves comparing the contour features with the corresponding features in the three-dimensional model using a feature point matching algorithm to achieve precise positioning. Adjustment and recalibration mechanism involves adjusting the preset processing parameters immediately upon detection of a deviation (this is the same principle as the benchmarking solution library mentioned in the model import and preprocessing module, so it will not be elaborated on here). A calibration notification is issued, prompting a secondary calibration action to complete the re-inspection or secondary adjustment to ensure that the final product meets the design requirements. For example, if the target workpiece is detected to have deviated by 0.5 mm, the corresponding solution in the solution library is immediately retrieved, i.e., the tool path is automatically adjusted to ensure that the processing accuracy is not affected. In an actual case, if a slight deviation of the target workpiece is detected during the processing of 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.

[0021] It should be noted that the machine vision recognition and calibration module greatly enhances the adaptability of the system, enabling it to maintain high processing accuracy even in the face of complex and changeable working conditions. Since the specific adjustment and re-inspection are not the key contents reflected in this embodiment and it is an existing solution, the adjustment and re-inspection calibration mechanism will not be elaborated in detail.

[0022] By adopting machine vision recognition and calibration technology solutions, precise positioning and dynamic adjustment of workpieces are achieved, which not only achieves the effect of real-time feedback, but also solves the inefficiency and inaccuracy problems caused by traditional manual correction. In traditional processing, the actual position of the workpiece after clamping often requires manual measurement and correction, which is not only time-consuming and labor-intensive, but also prone to human errors. This solution uses machine vision technology to capture the surface image of the workpiece in real time through a high-resolution camera network installed on the machine tool, and 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 found, the system will immediately adjust the processing parameters to ensure that the final product meets the design requirements, greatly improving processing efficiency and accuracy.

[0023] 3. Processing path planning module: Based on the accurate information provided by the first two modules, this module focuses on generating the optimal tool path; The specific implementation contents are: Based on the shape of the target workpiece and the machining target (i.e., the required machining area), several feasible tool paths are initially set. A simulation evaluation mechanism is performed on different tool path schemes to screen candidate paths from the several feasible tool paths. A secondary simulation evaluation mechanism is then triggered to extract the qualified path from the candidate paths. Among them, the setting of the preliminary path is based on the target workpiece geometry and the processing target. The system automatically generates several feasible tool paths by identifying the boundary contour, feature area (such as boss, groove) and processing depth; in this embodiment, the commonly used strategies in actual processing include: parallel cutting, circumferential cutting or reciprocating path to ensure that the entire processing area is covered; in actual application, for some more conventional rectangular workpieces, the system generates equidistant straight line paths along the X / Y direction based on the length and width dimensions and processing allowances, and adds transition arcs at the corners to ensure continuity and processing stability; each path consists of several cutting segments, and the system automatically marks key points (such as the starting point, reversal point, and cutting position) to facilitate subsequent optimization and simulation analysis.

[0024] The contents of a simulation evaluation mechanism executed are as follows: Suppose there is a set of paths P={P1, P2, ..., P n}; i = 1, 2, ..., n, n is a positive integer, i is the number of the corresponding path; Among them, each path P i It is necessary to calculate the processing time and material consumption; processing time T(P i ): The calculation formula for processing time is: ; Among them, m i is the path P i Number of cutting segments, t j is the time of the jth cutting segment, based on: t j =l j / v j ; l j is the length of the jth cutting segment, v j is the feed rate of the cutting section; material consumption M(P i ): The calculation formula for material consumption is: ; Among them, ml j is the material removal amount of the jth cutting section, based on: ml j =A j / d j Among them, A j is the area of the jth cutting, d j is the cutting depth of the segment; the first comprehensive evaluation function: the first comprehensive evaluation function F1(P i) is obtained by weighted summation based on processing time and material consumption: ; Among them, w1 and w2 are weight coefficients of processing time and material consumption, respectively, and their values range from 0 to 1; Screening path: Select F1(P i ) is smaller than the first set threshold as a number of candidate paths; wherein the first set threshold is set according to actual needs, usually screening out the paths that make F1(P i )The minimum or near-minimum path.

[0025] It should be noted that the result of executing the simulation evaluation mechanism above is: On the basis of selecting several paths that save time and reduce material waste, the tool wear and processing accuracy on different paths are further considered to ensure the balance between the two, so the subsequent mechanism needs to be implemented; The triggered secondary simulation evaluation mechanism content is: Tool wear W(P i ): The calculation formula for tool wear is: ; Among them, wl j is the wear of the tool in the jth cutting segment, estimated by the formula: wl j =k j ×l j ×d j ; where k j represents the wear coefficient of the jth cutting segment, l j represents the cutting length of the jth cutting segment, d j Indicates the cutting depth of the segment; machining accuracy Q(P i ): Calculation formula for machining accuracy: ;q j is the machining error of the jth cutting segment; it is estimated by the formula: q j =e j ×l j ×d j , where e j represents the error coefficient of the jth cutting segment, l j Indicates the cutting length of the segment, d j Indicates the cutting depth of the cutting segment; The second comprehensive evaluation function: The second comprehensive evaluation function F2 (P i ) is obtained by weighted summation based on tool wear and machining accuracy: ; Among them, w3 and w4 are the weight coefficients of tool wear and processing accuracy respectively, and the value range is 0 to 1; Second screening path: select several candidate paths that make F2(Pi) less than the second set threshold as the qualified path; Among them, the second set threshold and the first set threshold are also set according to actual needs, and usually the path that makes F2(Pi) minimum or close to minimum is screened out.

[0026] By adopting multi-level simulation and path optimization technology solutions, the optimal path is selected, which not only achieves the effect of reducing material waste and improving processing efficiency, but also solves the problem of difficult balance between time and cost in traditional path planning; traditional path planning methods usually rely on experienced technicians to manually set the path, or set it based on historical established paths, which makes it difficult to strike a balance between processing efficiency and material consumption; this solution first generates a preliminary tool path based on the target workpiece shape, and then uses primary and secondary simulation evaluation mechanisms to screen out paths that save time and reduce material waste; on this basis, further consideration is given to tool wear and processing accuracy, and the path with minimal wear and highest processing accuracy is selected for processing, thereby achieving the purpose of double balance and effectively improving overall processing efficiency and quality.

[0027] 4. Virtual simulation and optimization module: After completing the path planning, the system will enter the simulation stage, 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; The simulation test strategy executed is as follows: S101, Environment Construction: Build a highly realistic virtual machining environment, including all components such as machine tools, cutting tools, and fixtures, which need to be accurately modeled; Figure 4 It can be seen that precise instructions are given on material standardization, material digitization, and programming standardization; S102, process simulation: Performing a comprehensive simulation run on each path to meet the standard, detecting whether any cutting segment in the path to meet the standard has any abnormality, and performing corresponding marking operations on the detection results; Among them, the abnormal phenomenon is overcutting or collision; each standard path includes several path segments, and each path segment is composed of several cutting segments; for example, if a standard path is in the shape of a square, then there are four path segments, and the cutting segments of each path segment can be set to five; because the path segment replacement operation needs to be introduced later, the original standard path needs to be replaced; For details, please refer to the attached Figure 3 As shown in the figure, the system interface is for the path group formed by the target path: The corresponding cutting segments with overcutting or overcutting and collision are displayed in red; the corresponding cutting segments with collision are displayed in yellow; the corresponding cutting segments without overcutting or collision are displayed in green. In this embodiment, after a large number of simulations, it was found that there is a significant correlation between the red and yellow density and the subsequent tool wear and processing accuracy. Therefore, the red and yellow density is not only a visual abnormality mark, but also an important indicator for predicting processing performance.

[0028] S103, extracting the path with the largest proportion of green marks among the paths that meet the standards as the preliminary path; S104. For the initially determined 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, use the initially determined path as the final path; If a high-risk segment is identified, the path optimization mechanism is triggered: the system returns to the previous level and searches for alternative path segments from the qualified path based on preset standard conditions to replace the current high-risk segment. After obtaining the other path segments that meet the conditions, they are inserted into the original path to generate a new path, and the second comprehensive evaluation function of the new path is recalculated. The second comprehensive evaluation functions of the original path and the optimized new path are compared, and the smaller one is selected as the final path. Among them, for the initial path, traverse all cutting segments S in its path segment j , count the number of red and yellow cutting segments, and calculate their proportion in each path segment in the initial path: ; Among them, N red (P b ), N yellow (P b ) and N total (P b ) respectively represent: in any path segment of the initial path, the number of red marking cutting segments, the number of yellow marking cutting segments and the total number of cutting segments; if D ry (P b )>The set threshold T corresponding to the density of the red and yellow segments ry , 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 default standard conditions indicate: Geometric position matching: the replacement path segment should be in the same or similar geometric area as the original path segment; Functional consistency: cutting type (roughing / finishing), direction and depth parameters are the same or similar; Low non-green cutting density: the density of non-green cutting segments in the replacement path segment should be less than 1 / 2 of the original path segment; Among them, the parameters are close, which means that the corresponding parameters are within the error range; Alternative path segments are screened based on three criteria: geometric position matching, functional consistency, and low non-green density, ensuring that the replaced paths meet functional, structural, and safety requirements. During actual operation, the system discovered that some paths that were excluded due to poor overall performance had segments with extremely high replacement value. Therefore, this embodiment uses the above-mentioned preset standard conditions for screening, no longer pursuing perfect paths, but instead focusing on the reuse potential of path segments. This solution breaks with this inertial thinking and proposes the concept of "prioritizing path segment availability." This reconstructs path resource management from a micro perspective, achieving breakthroughs and optimizations in current technological barriers.

[0029] Compare the second comprehensive evaluation function of the original path P0 (i.e., the initial path) and the optimized path P0', and select the one with the smaller value 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. By dividing the qualified path into multiple path segments and marking 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, so that the path that may have local defects is better than the original complete path after local replacement. It should be noted that traditional path optimization mostly adopts a global replacement strategy, believing that the path as a whole cannot be split, and replacing the entire path is the current mainstream practice. In the solution given in this embodiment, dynamic replacement based on path segments is proposed for the first time, and combined with a secondary evaluation mechanism of a comprehensive evaluation function, the path quality is improved 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.

[0030] The advantages of implementing the path optimization mechanism in the simulation test strategy are as follows: 1. Reasonable division of path segments: ensuring that each path segment is independent and replaceable; 2. Strict screening of alternative segments: geometric position, cutting function, and red and yellow density must all match; 3. Comprehensive performance evaluation: comprehensive consideration of tool wear and processing accuracy; 4. High degree of automation: the entire process can be automatically executed by the system, and the actions or operations to confirm the alternative strategy can be automatically completed; in addition, machine learning algorithms can be introduced based on this mechanism, and the model can be trained based on historical data to automatically predict which path segments are more likely to have red and yellow problems, so as to avoid these areas in the path planning stage, realize active risk-avoidance path optimization, and further improve the level of intelligent manufacturing.

[0031] The final path is then optimized twice to generate the optimal path; The secondary optimization process includes reducing the feed rate, changing the cutting direction, increasing the coolant injection, and using an adaptive control algorithm. Under the conditions of the simulation stage: Reduce feed rate: In the cutting section where overcutting or collision occurs, reduce the feed rate appropriately to reduce the impact of the tool on the workpiece; Example: If an overcutting problem is detected in a certain cutting section, the system reduces the feed rate of the section from the original 500mm / min to 300mm / min. mm / min, after re-simulation, it was found that the overcutting problem disappeared; change the cutting direction: change the cutting direction of the tool to avoid collision with the target workpiece; example: suppose that in the process of processing a complex surface, there is a collision problem in a certain path segment. The system changes the cutting direction of the path segment from clockwise to counterclockwise. After re-simulation, it was found that the collision problem disappeared; increase coolant injection: in high-temperature areas or areas prone to overcutting, add coolant injection function to reduce temperature and reduce overcutting; example: suppose that when processing titanium alloy blades, there is an overcutting problem in a certain path segment. The system adds a coolant injection function to this path segment. After re-simulation, it is found that the overcutting problem disappears; adaptive control algorithm: use the adaptive control algorithm to dynamically adjust the tool's motion trajectory according to the changes in the target workpiece surface to ensure processing quality and efficiency; example: suppose that when processing a complex surface, there is an overcutting problem in a certain path segment. The system uses an adaptive control algorithm to dynamically adjust the tool's motion trajectory according to the changes in the surface. After re-simulation, it is found that the overcutting problem disappears.

[0032] By adopting comprehensive simulation testing and path optimization technology solutions, closed-loop verification of the processing path is achieved, which not only prevents potential risks but also solves the problems of frequent rework and resource waste in traditional trial cutting processes. In traditional machining processes, due to the lack of effective simulation verification methods, multiple trial cuts are often required to find the right machining path, resulting in resource waste and extended production cycles. This solution builds a highly realistic virtual machining environment to conduct comprehensive simulation tests on each path to check for overcutting or collisions. For problematic path segments, the system automatically searches for alternative path segments to replace them, and performs a secondary comparison with the original path after replacement to screen out the optimal path. This can reduce the amount of secondary optimization processing operations to a certain extent, improving machining safety and stability on the one hand, and enhancing work efficiency to a certain extent on the other.

[0033] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will appreciate 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 performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0034] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, and may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment as needed.

[0035] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A simulation virtual processing system based on machine vision, characterized in that: include: The model import and preprocessing module imports the 3D model of the target workpiece, performs format conversion processing, detects potential problem sources, compares the repair solutions given in the solution library, completes the repair process, and completes data standardization processing; The machine vision recognition and calibration module uses machine vision technology to locate, capture, and match features of the target workpiece. When deviations are identified, an adjustment and recalibration mechanism is triggered to ensure that the target workpiece meets the design requirements. The machining path planning module preliminarily sets several tool paths based on the shape of the target workpiece and the machining objectives. It then performs a primary simulation evaluation mechanism on different tool path schemes to screen out candidate paths and triggers a secondary simulation evaluation mechanism to extract the qualified path from the candidate paths. 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.

2. The machine vision-based simulation virtual machining system according to claim 1, characterized in that: When performing format conversion processing, multiple file formats are supported, including 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 repair solutions given by the benchmark solution library include at least: non-manifold edge repair solution, self-intersecting surface repair solution and geometric topology repair solution; data standardization processing: the imported 3D model is unified in size unit.

3. The machine vision-based simulation virtual machining system according to claim 1, characterized in that: The process of locating and capturing the target workpiece is as follows: using a pre-installed visual 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: using a feature point matching algorithm 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 machining system according to claim 3, characterized in that: The triggered adjustment and re-inspection calibration mechanism process is as follows: when a deviation is identified, the preset processing parameters are adjusted, and a calibration notification is issued, prompting a secondary calibration action to complete the re-inspection or secondary adjustment.

5. The machine vision-based simulation virtual machining system according to claim 1, characterized in that: The preliminary path setting is based on the target workpiece geometry and machining objectives, and several tool paths are generated by identifying boundary contours, feature areas, and machining depths. Among them, the strategies adopted during processing include at least: parallel cutting, circumferential cutting and reciprocating path.

6. The machine vision-based simulation virtual machining 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, ..., P n }; Each path P i It is necessary to calculate its processing time and material consumption; wherein, the processing time is obtained based on the number of cutting segments and the time of each cutting segment, and the time of each cutting segment is generated based on the length of each cutting segment and its corresponding feed speed; the material consumption is obtained based on the number of cutting segments and the material removal amount of each cutting segment, and the material removal amount of each cutting segment is generated based on the area of each cutting segment and its corresponding cutting depth; the first comprehensive evaluation function is obtained by weighted summation of the processing time and material consumption; several paths whose first comprehensive evaluation function is less than the first set threshold are screened out and used as candidate paths.

7. The machine vision-based simulation virtual machining system according to claim 6, characterized in that: The triggered secondary simulation evaluation mechanism content is: The tool wear is calculated based on the number of cutting segments and the amount of tool wear caused by each cutting segment. The amount of tool wear caused by each cutting segment is calculated based on the wear coefficient, cutting length and cutting depth of each cutting segment. The 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 generated 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 taken as qualified paths.

8. The machine vision-based simulation virtual machining system according to claim 1, characterized in that: The simulation test strategy process executed is: Environment construction: Build a virtual machining environment, including at least: machine tools, cutting tools and fixtures; Process simulation: A comprehensive simulation is performed on each path to detect any anomalies in any cutting segment along the path, and corresponding identification is performed on the detection results. Anomalies include overcutting or collisions. Each path consists of several path segments, and each path segment is composed of several cutting segments. The corresponding cutting segments with overcutting, overcutting and collision are displayed in red; The corresponding cutting segments where there is a collision are displayed in yellow; The corresponding cutting segments without overcutting or collision are displayed in green; Result analysis: The path with the largest proportion of green markings among the qualified paths is extracted as the preliminary path. For the preliminary 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 preliminary path is used as the final path. Otherwise, the path optimization mechanism is triggered.

9. The machine vision-based simulation virtual machining system according to claim 8, characterized in that: The triggered path optimization mechanism is as follows: according to the preset standard conditions, other path segments that can replace the current high-risk segment are found from the qualified path; after obtaining other path segments that meet the conditions, they are inserted into the original path to generate a new path, and its second comprehensive evaluation function is recalculated; the second comprehensive evaluation function of the original path and the optimized new path are compared, and the smaller one is selected as the final path.

10. The machine vision-based simulation virtual machining system according to claim 8, characterized in that: For the initial 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 in the initial path ry (P b ), if D ry (P b )> 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 indicate: 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 the original path segment.

Images