Method of combined path and laser machining planning based on machine vision

Abstract

The present application relates to the technical field of path control, in particular to a method for combined path and laser processing planning based on machine vision, comprising the following steps: an industrial camera acquires an image, extracts a profile, measures parameters, identifies defects and materials, sets a path start and end based on a feature table, optimizes connection and smoothing, avoids obstacles, detects a hot area, dynamically adjusts a range, sets a threshold based on an adaptive group, evaluates ability, calibrates delay, predicts adjustment power and speed, calls an instruction set, compensates for prediction, corrects errors, dynamically matches, and outputs an optimal processing path and a synchronous processing setting table. In the present application, by acquiring workpiece images in real time and analyzing processing features, the processing precision and efficiency are improved, fine control during processing is realized, effective cooperation of path planning and processing parameters is ensured, path planning mismatching problems are avoided, response accuracy and stability during processing are improved, processing quality and time management are significantly improved, and the low efficiency and quality fluctuations caused by manual adjustment in traditional technology are avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of path control, in particular to a method for combined path and laser processing planning based on machine vision. BACKGROUND

[0002] The technical field of laser processing involves the interaction of laser beams with materials to achieve high-precision processing. This field mainly includes laser cutting, laser marking, laser welding, and other processes. Path control is one of the core technologies, focusing on how to optimize and plan laser paths according to processing tasks to ensure efficient and accurate processing. The goal of path planning is to maximize processing precision and minimize processing time while addressing the complexity of different materials. This field also includes laser power regulation and processing speed control to ensure stability and efficiency during processing. Traditional laser processing path control methods involve controlling the laser processing process through pre-set fixed paths. However, in practical applications, there are often problems such as poor coordination between path planning and processing parameters, and poor adaptability to complex situations. Traditional techniques rely on simple path planning algorithms and manual adjustment of processing parameters, ignoring changes in actual working conditions during processing, resulting in low efficiency or unstable processing quality.

[0003] Existing path control methods rely on pre-set paths and ignore the dynamic changes in actual working conditions during processing, leading to insufficient coordination between path planning and processing parameters. In practical applications, traditional techniques cannot adjust in real-time according to the actual shape, surface defects, or material properties of the workpiece, which can lead to low processing precision or long processing time. Due to the use of fixed path planning algorithms and manual adjustment of processing parameters, complex working conditions cannot be addressed, resulting in low efficiency or unstable quality. For example, in the processing of complex-shaped workpieces, the rigidity of path planning and the lack of real-time data feedback adjustment mechanism make the path unable to adapt to the curvature or surface changes of the workpiece, resulting in wasted processing time or poor product surface quality. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and to provide a method for combined path and laser processing planning based on machine vision.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solution: a method for combined path and laser processing planning based on machine vision, comprising the following steps:

[0006] S1: Real-time acquisition of processing area images through an industrial camera, contour extraction and geometric parameter measurement of the workpiece shape, classification and spatial positioning of workpiece surface defects, physicochemical property identification and component composition analysis of the workpiece material, and generation of a processing feature distribution table;

[0007] S2: based on the machining feature distribution table, the starting point and the ending point of the multi-objective machining path are set, the connection order optimization of the multi-path segment is performed, the smoothing processing of the path inflection point and the path of the obstacle avoidance area are adjusted, and a path planning priority list is generated;

[0008] S3: according to the path planning priority list, the real-time measurement and change trend capture of the workpiece surface curvature are performed, the boundary dynamic detection and diffusion evaluation of the machining area and the heat affected zone are performed, the range demarcation and condition matching of the machining interval are dynamically adjusted, and a machining adaptation relationship group is established;

[0009] S4: according to the machining adaptation relationship group, the parameter threshold setting and execution capability evaluation of the current machining requirement are performed, the delay time and stability prediction of the response capability are calibrated, the continuous stepless adjustment of the laser power and the synchronous adjustment of the scanning speed are performed, and a machining parameter cooperative control instruction set is output.

[0010] As a further scheme of the application, the machining feature distribution table comprises workpiece number, shape feature identification, defect distribution mark, material property label, the path planning priority list comprises path weight value, machining sequence identification, priority level, the machining adaptation relationship group comprises curvature change interval, heat affected range identification, machining feasibility determination value, and the machining parameter cooperative control instruction set comprises power level code, speed adjustment direction, and periodical adjustment identification.

[0011] As a further scheme of the application, the generation step of the machining feature distribution table is specifically:

[0012] S111: real-time acquisition of machining area images is performed through an industrial camera, the images are input into a deep learning model for processing, pixel-level features are extracted and mapped, workpiece contour boundaries are identified, topological structures are analyzed, surface defect categories are identified and spatial position positioning is performed, material spectral response characteristics are analyzed and physical properties are inferred, and a machining feature sequence is generated;

[0013] S112: the machining feature sequence is called to perform matching analysis of workpiece geometric morphology and defect distribution mode, risk defect areas are identified and machining priority is labeled, and a key machining area index set is generated;

[0014] S113: according to the key machining area index set, geometric shape consistency areas are identified, material uniformity areas are verified, non-critical defect areas are marked and machining stability is evaluated, and a machining feature distribution table is generated.

[0015] As a further scheme of the application, the generation step of the path planning priority list is specifically:

[0016] S211: Call the machining feature distribution table, extract the geometric feature and boundary information of the key machining area, identify the defect type, size, position information and carry out quantization, carry out preliminary construction and key node setting of multiple alternative paths, evaluate and convert the numerical value of path length, bending degree, obstacle avoidance, surface roughness requirement factors, and generate path weight set;

[0017] S212: According to the path weight set, compare and analyze the weight value of each path and judge the size, arrange the path, determine the priority level of the path, and generate the path priority sequence;

[0018] S213: Based on the path priority sequence, the compliance of each path with the preset machining parameter is verified, and the path that does not meet the machining precision requirement, has material damage risk or low machining efficiency is identified and filtered, and the path planning priority list is generated.

[0019] As a further scheme of the application, the generation step of the machining adaptation relationship group is specifically:

[0020] S311: Call the path planning priority list, carry out real-time scanning and three-dimensional reconstruction of the workpiece surface curvature, identify the curvature gradient and mutation point, dynamically monitor the machining area temperature field and construct the heat diffusion model, track and evaluate the morphology of the heat affected area boundary in real time, and obtain the machining environment state parameter by summarizing the curvature change and heat affected data;

[0021] S312: According to the machining environment state parameter, the upper and lower limits of the curvature change range are set, and the allowable threshold of the heat affected range is defined and the safety boundary is drawn, the region parameter data is compared and analyzed point by point, the region that does not meet the machining precision or material integrity requirement is identified and removed, the interval that meets the machining requirement is screened, and the machining available interval set is obtained;

[0022] S313: Call the machining available interval set, carry out spatial positioning and geometric matching of the path segment in the machining path planning priority list, carry out overlap calculation and compatibility evaluation of the path segment and the available interval, analyze the corresponding relationship of the path and the interval characteristics, identify the machining parameter adjustment potential, map the adaptation relationship between the machining interval and the path, calculate the evaluation index value of the machining adaptation relationship, and generate the machining adaptation relationship group.

[0023] As a further scheme of the application, the generation step of the machining parameter cooperative control instruction set is specifically:

[0024] S411: According to the machining adaptation relationship group, the surface roughness requirement and the topography precision index in the current machining interval are extracted, the material removal rate target and the thermal damage control threshold are set, the laser power adjustment range and the response speed are quantitatively evaluated, the scanning galvanometer speed and acceleration performance are detected, the machining demand and the response capability are matched, and the machining demand deviation data group is generated;

[0025] S412: The machining demand deviation data group is called, the parameter deviation value is compared with the preset allowable range one by one, the deviation items exceeding the allowable range are identified and classified, the matching degree between the machining demand and the response capability is hierarchically divided and state represented, the key parameter matching degree insufficient area is marked, and the matching state identification set is generated.

[0026] S413: According to the matching state identification set, the retrieval and spatial position of the machining interval number are verified, the increment or decrement of the required laser power in the current interval is calculated, the acceleration or deceleration amplitude of the scanning speed is set, the power and speed adjustment instructions are identified, and the machining parameter cooperative control instruction set is generated.

[0027] As a further scheme of the application, the quantitative evaluation of the laser power adjustment range and the response speed refers to detecting and quantifying the dynamic adjustment capability of the laser power and the response speed of the laser to the adjustment instruction.

[0028] The one-by-one comparison of the parameter deviation value and the preset allowable range refers to comparing the laser power adjustment range deviation, the response speed deviation, the scanning galvanometer speed deviation and the scanning galvanometer acceleration deviation included in the machining demand deviation data group with the preset allowable range respectively.

[0029] As a further scheme of the application, the method further comprises a step S5:

[0030] S5: The machining parameter cooperative control instruction set is called, real-time delay compensation and prediction of the machining path are performed, fluctuation characteristics and error correction of the response time are analyzed, dynamic correction of the path and corresponding machining parameters are performed, and the optimal machining path and the synchronous machining setting table are output.

[0031] The optimal machining path and the synchronous machining setting table comprise path delay characteristics, synchronous machining time points and response characteristics.

[0032] As a further scheme of the application, the generation step of the optimal machining path and the synchronous machining setting table is specifically:

[0033] S511: Call the processing parameter cooperative control instruction set, analyze the difference between the laser scanning head motion instruction and the real-time response time sequence, perform power adjustment and scanning galvanometer cooperative action lag monitoring, measure the fluctuation range of the response time under differentiated load conditions, calculate the path delay fluctuation characteristic value, and generate the path delay fluctuation correlation coefficient;

[0034] S512: According to the path delay fluctuation correlation coefficient, analyze the complexity of the path geometry, and perform parameter mapping of material thickness and hardness change, set the precision requirement and stability margin of synchronous processing conditions, evaluate the continuity and smoothness constraint between path segments, select the path set that meets the synchronous processing conditions, and generate the synchronous processing decision value interval;

[0035] S513: Call the synchronous processing decision value interval, perform response speed and precision matching of laser power adjustment, and perform acceleration limitation and smoothness comparison of scanning galvanometer speed change, optimize the transition time and seamless connection of parameter switching between path segments, determine the optimal processing path and corresponding processing parameters, and output the optimal processing path and synchronous processing setting table.

[0036] Compared with the prior art, the advantages and positive effects of the present application are:

[0037] In the present application, by real-time acquisition of workpiece image and processing feature analysis, the laser processing path is dynamically optimized based on the actual situation of the workpiece, improving the accuracy and efficiency of the processing process. Through automatic extraction of the geometric parameters of the workpiece, defect classification and material identification, fine control in the processing process is realized, ensuring the effective cooperation of path planning and processing parameters. Dynamically adjusting the range and conditions of the processing area, optimizing the path and parameters in the processing process according to real-time data such as surface curvature and heat-affected zone, avoids the problem of mismatched path planning in processing. The synchronous adjustment of laser power and scanning speed makes the response capability in the processing process more accurate and stable, significantly improving the processing quality and time management, thereby avoiding the low efficiency and quality fluctuation caused by manual adjustment in traditional technology. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is the main step schematic diagram of the present application;

[0039] Figure 2 is the acquisition flowchart of the processing feature distribution table in the present application;

[0040] Figure 3 is the acquisition flowchart of the path planning priority list in the present application;

[0041] Figure 4 is the acquisition flowchart of the processing adaptation relationship group in the present application;

[0042] Figure 5 This is a flowchart illustrating the acquisition of the collaborative control instruction set for processing parameters in this invention.

[0043] Figure 6 This is a flowchart illustrating the process of obtaining the optimal processing path and synchronous processing setting table in this invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0045] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Example 1

[0046] Please see Figure 1 This invention provides a technical solution: a method for planning combined paths and laser processing based on machine vision, comprising the following steps:

[0047] S1: Real-time acquisition of images of the processing area through an industrial camera, extraction of the workpiece shape contour and measurement of geometric parameters, classification and spatial positioning of workpiece surface defects, identification of the physical and chemical properties of the workpiece material and analysis of its composition, and generation of a processing feature distribution table.

[0048] S2: Based on the processing feature distribution table, set the start and end points of the multi-target processing path, optimize the connection order of multiple path segments, adjust the smoothing of path inflection points and the path in the obstacle avoidance area, and generate a path planning priority list.

[0049] S3: Based on the path planning priority list, perform real-time measurement and trend capture of the workpiece surface curvature, and perform dynamic detection and diffusion assessment of the boundaries of the processing area and heat-affected area, dynamically adjust the range delineation and condition matching of the processing area, and establish a processing adaptation relationship group.

[0050] S4: According to the processing adaptation relationship group, the parameter threshold setting and execution capability assessment of the current processing demand are performed, and the delay time and stability prediction of the response capability are calibrated to output the processing parameter cooperative control instruction set for continuous stepless adjustment of the laser power and synchronous adjustment of the scanning speed.

[0051] S5: The processing parameter cooperative control instruction set is called to perform real-time delay compensation and prediction of the processing path, analyze the fluctuation characteristics and error correction of the response time, perform dynamic correction of the path and corresponding processing parameters, and output the optimal processing path and synchronous processing setting table.

[0052] The processing feature distribution table includes workpiece number, shape feature identifier, defect distribution label, and material property label. The path planning priority list includes path weight value, processing sequence identifier, and priority level. The processing adaptation relationship group includes curvature change interval, heat affected range identifier, and processing feasibility determination value. The processing parameter cooperative control instruction set includes power level code, speed adjustment direction, and periodical adjustment identifier. The optimal processing path and synchronous processing setting table includes path delay characteristics, synchronous processing time point, and response characteristics.

[0053] Please refer to Figure 2 The generation step of the processing feature distribution table is specifically as follows:

[0054] S111: Real-time acquisition of the processing area image is performed through an industrial camera, the image is input into a deep learning model for processing, pixel-level feature extraction and mapping are performed, the workpiece contour boundary is identified, the topological structure is analyzed, the category identification and spatial position positioning of surface defects are performed, the material spectral response characteristics are analyzed and the physical properties are inferred, and the processing feature sequence is generated.

[0055] A workpiece surface image stream is captured in real time through an industrial camera, the resolution is 2560x2048, the frame rate is 30 frames / second, the vertical shooting is performed, the image data is high-speed transmitted to an image processing unit, and a deep learning model is input. The model is based on a U-Net architecture, the encoder extracts multi-scale features such as edges and textures, the decoder fuses high-level semantics and low-level spatial details, generates a pixel-level feature vector, and encodes the pixel region type. The workpiece contour boundary is identified, the pixels are classified as “workpiece contour” or “background” through a binary classifier, the contour pixels are connected through morphological operation, the closed contour is obtained, the topological structure is analyzed, the Euler number is calculated, and the number of holes and boundaries is identified. The surface defect category identification and spatial positioning are performed, the model outputs the defect category, mask and bounding box coordinates, a multi-spectral imaging system obtains reflectance spectral data to form a spectral vector, a neural network model is input, the model infers physical properties such as material hardness and thermal conductivity, and a processing feature sequence is generated.

[0056] S112: Call the machining feature sequence, perform matching analysis of workpiece geometry and defect distribution pattern, identify risk defect area and mark machining priority, generate key machining area index set;

[0057] Call the machining feature sequence, extract workpiece contour point set, defect type and position list, material physical property data, perform matching analysis of workpiece geometry and defect distribution pattern, discretize the workpiece contour point set into a series of geometric features, for example, calculate the area, perimeter, aspect ratio, and barycenter coordinates of the workpiece, and compare them with the preset standard geometric shape template, for example, compare the aspect ratio of the workpiece with the ratio of the standard rectangular template, at the same time, analyze the defect distribution pattern, for example, calculate the defect density (number of defects per unit area), average distance between defects, and degree of defect clustering, identify risk defect area and mark machining priority, integrate the defect type, size, and position information with the workpiece geometry matching analysis results, identify the defect area that has a large impact on subsequent machining quality as the risk defect area, for example, cracks near the workpiece boundary (areas within 2 mm of the boundary) are identified as high-risk defects, with a priority mark of "1", single small pores (diameter less than 0.2 mm) inside are marked as low-risk, with a priority of "3", the priority division rules are as follows: priority 1 represents "high", when the defect type is crack, and the defect size (length or diameter) is greater than 1% of the workpiece feature size, or the defect is located in the key geometric feature area; priority 2 represents "medium", when the defect type is scratch or surface pit, and the defect size is between 0.1% and 1% of the workpiece feature size; priority 3 represents "low", when the defect type is non-metallic inclusion or slight oxidation layer, and the defect size is less than 0.1% of the workpiece feature size, generate key machining area index set.

[0058] S113: According to the key machining area index set, identify geometric shape consistency area, verify material uniformity area, mark non-critical defect area and evaluate machining stability, generate machining feature distribution table;

[0059] According to the key machining area index set, identify geometric shape consistency area, analyze local curvature of workpiece contour point set, evaluate local smoothness and consistency, verify material uniformity area, use material physical property distribution data to analyze hardness value standard deviation, determine material uniformity area, mark non-critical defect area, evaluate machining stability, identify defects with priority 3, mark defects as non-critical, analyze material uniformity, geometric consistency and preset machining parameters around the defects, evaluate regional machining stability, the stability evaluation value is calculated by weighting the geometric consistency score, material uniformity score and defect severity, evaluate value high is determined as "high" stability, medium is determined as "medium", low is determined as "low", generate machining feature distribution table.

[0060] Referring to Figure 3 The path planning priority list generation step is specifically as follows:

[0061] S211: Call the machining feature distribution table, extract the geometric feature and boundary information of the key machining area, identify the defect type, size, and position information and perform quantization, perform preliminary construction of multiple alternative paths and key node setting, evaluate and convert the numerical values of path length, bending degree, obstacle avoidance, and surface roughness requirement factors, and generate a path weight set;

[0062] Call the machining feature distribution table, including key machining area index, defect type, size, position information, and machining stability evaluation result structured data, extract key machining area geometric shape features and boundary information, for each key machining area, extract the accurate boundary coordinate point set, and calculate the geometric shape features, for example, calculate the minimum circumscribed rectangle, center point coordinates, area, perimeter, and its relative position to the overall geometric center of the workpiece, for example, a high-risk crack area is marked, the boundary information is represented by a closed polygon point set, the area is 1.2 square millimeters, the defect type, size, and position information are identified and quantified, all defect information in the key machining area is arranged, for example, the defect types include cracks, scratches, and pores, the defect sizes include crack length, scratch depth, and pore diameter, the defect position is given by the (x, y) coordinates relative to the workpiece coordinate system origin (0, 0), and the information is quantified, for example, crack length 5.2 mm, scratch depth 0.05 mm, and pore diameter 0.3 mm, perform preliminary construction of multiple alternative paths and key node setting, based on the geometric shape features of the key machining area and the defect position, generate multiple laser scanning paths, the paths can be straight lines, arcs, or compound curves, each path contains multiple key nodes, for example, path 1 is [(10, 10), (20, 20), (30, 30)], path 2 is [(10, 10), (15, 15), (25, 25), (30, 30)], and path 3 is [(10, 10), (20, 15), (30, 10)], the nodes define the path direction and turning points, evaluate and convert the path length, curvature, obstacle avoidance, and surface roughness requirement factors to numerical values, each alternative path is evaluated in multiple dimensions, the path length is directly calculated as the sum of the lengths of all line segments on the path, for example, the length of path 1 is 28.28 mm, the curvature is measured by calculating the sum of the deviations of the angles at each turning point from 180 degrees, the obstacle avoidance evaluation assesses whether the path interferes with non-machining areas or already identified non-critical defect areas on the workpiece, if interference occurs, the obstacle avoidance score is 0, otherwise it is 1, the surface roughness requirement is assigned according to the specific requirements of the target machining area, for example, for a high-precision surface, the roughness requirement is Ra less than 0.8 microns, and the assignment is 10, and a path weight set is generated. The weight values in the path weight set are obtained by weighted summation of the above evaluation factors, for example, the path weight is calculated by the normalized values of length, curvature, obstacle avoidance, and surface roughness requirement and their corresponding weight coefficients, the weight coefficients are set according to actual machining requirements, for example, length weight 0.3, curvature weight 0.2, obstacle avoidance weight 0.4, and surface roughness requirement weight 0.1.

[0063] S212: According to the path weight set, compare and analyze the weight values of each path, and determine the size, arrange the paths, determine the priority level of the paths, and generate a path priority sequence;

[0064] According to the path weight set, S calculates the path weight value of each candidate path, for example, path A weight 0.75, path B weight 0.82, path C weight 0.69, compares and judges the size of each path weight value, compares each path weight value with each other to determine the relative size relationship, for example, path B weight 0.82 is greater than path A weight 0.75, path A weight 0.75 is greater than path C weight 0.69, arranges the path weight values from large to small through the sorting algorithm, arranges the paths, arranges the paths in descending order according to the size relationship of the weight values, forms an ordered path list, for example, path B (0.82), path A (0.75), path C (0.69), determines the path priority level, assigns a priority level to each path according to the arrangement order, for example, the first path B obtains priority 1, the second path A obtains priority 2, and the third path C obtains priority 3, and generates a path priority sequence.

[0065] S213: Based on the path priority sequence, the compliance of each path with the preset machining parameters is verified, and the paths that do not meet the machining precision requirements, have material damage risk or low machining efficiency are identified and filtered to generate a path planning priority list;

[0066] Based on the path priority sequence, according to the ordered path list and the corresponding priority level, the compliance of each path with the preset machining parameters is verified, and the paths that do not meet the machining precision requirements, have material damage risk or low machining efficiency are identified and filtered to generate a path planning priority list;

[0067] Please refer to Figure 4 , the generation step of the machining adaptation relationship group is specifically:

[0068] S311: Call the path planning priority list, perform real-time scanning and three-dimensional reconstruction of the workpiece surface curvature, identify curvature gradient and mutation points, dynamically monitor the processing area temperature field and construct a thermal diffusion model, track and evaluate the morphology of the thermal influence area boundary in real time, and obtain the processing environment state parameters by summarizing the curvature change and thermal influence data;

[0069] Call the path planning priority list, including screening and priority sorting of the candidate path list, perform real-time scanning and three-dimensional reconstruction of the workpiece surface curvature, use a confocal laser scanner or a structured light scanner to perform real-time three-dimensional morphology scanning on the workpiece processing area, for example, collect surface point cloud data of the processing area at a speed of 20,000 points per second through laser triangulation principle, and use the iterative closest point algorithm to register and fuse the continuously collected point cloud data to construct a three-dimensional model of the processing area, for example, reconstruct a three-dimensional grid model containing 1 million vertices, identify curvature gradient and mutation points, calculate the surface normal vector at each grid vertex, and calculate the included angle between adjacent vertex normal vectors to obtain the local curvature value of the surface, then calculate the rate of change of the curvature value along the surface direction, for example, if the regional curvature gradient exceeds 0.5, the region is identified as a curvature mutation point, and the mutation point corresponds to an edge, a corner or a hole edge, dynamically monitor the processing area temperature field and construct a thermal diffusion model, use an infrared thermal imager to monitor the real-time temperature of the laser processing area, for example, obtain a real-time temperature distribution map of the processing point at a frame rate of 30Hz, with a data resolution of 640x480 pixels, then based on the real-time temperature data and the thermal physical parameters of the material, construct a transient thermal diffusion model, solve the heat conduction equation in the model, for example, the thermal conductivity of the material is 15 W / (m·K), the specific heat capacity is 450 J / (kg·K), and the density is 7900 , the model predicts the evolution of the internal and surface temperature field of the material under the action of the laser, tracks and evaluates the morphology of the thermal influence area boundary in real time, identifies and tracks the region where the temperature of the material exceeds the recrystallization temperature, for example, 850 degrees Celsius for 304 stainless steel, which is the heat-affected zone, extracts the boundary of the heat-affected zone through image processing algorithm, and evaluates the morphology characteristics such as the width, depth and whether it causes surface bulging or depression, and integrates the curvature change and thermal influence data to obtain the processing environment state parameters.

[0070] S312: According to the processing environment state parameters, set the upper and lower limits of the curvature change range, define the allowable threshold of the thermal influence range and demarcate the safety boundary, compare and analyze the region parameter data point by point, identify and exclude the regions that do not meet the processing precision or material integrity requirements, screen the intervals that meet the processing requirements, and obtain the processing available interval set;

[0071] According to the processing environment state parameters, the curvature gradient, curvature abrupt change point, heat affected zone boundary and topography data are summarized, the upper and lower limits of the curvature change range are set, based on the mechanical properties of the workpiece material and the expected processing precision requirements, the allowable range of surface curvature change is set, for example, for brittle materials or high-precision processing, the upper limit of the curvature change rate is set to 0.2, and the lower limit is set to 0.01, and the curvature change outside this range will cause processing difficulties or precision decline, and the heat affected range tolerance threshold is defined and the safety boundary is drawn, according to the material melting point, phase transition temperature and the requirement of avoiding thermal damage, the maximum allowable width and depth of the heat affected zone is set, for example, for laser welding of 304 stainless steel, the heat affected zone width tolerance threshold is defined as 0.5mm, and the depth tolerance threshold is 0.3mm, exceeding the threshold value will have the risk of thermal damage, at the same time, a safety boundary is drawn at least 1mm away from the sensitive area on the workpiece to prevent secondary damage caused by heat diffusion, the region parameter data is compared point by point and deviation analysis is performed, the real-time curvature change rate of each point in the processing area is compared with the set curvature change range, and the real-time width of the heat affected zone is compared with the tolerance threshold, for example, the real-time curvature change rate of a certain point is 0.25, which exceeds the set upper limit of 0.2, and the point is marked as a curvature unqualified point, the real-time width of the heat affected zone is 0.6mm, which exceeds the 0.5mm tolerance threshold, and the point is marked as a heat affected unqualified point, the areas that do not meet the processing precision or material integrity requirements are identified and removed, according to the comparison results, all areas that are not suitable for processing due to curvature change exceeding the range or heat affected exceeding the tolerance threshold are identified and removed from the areas to be processed, for example, all areas where the curvature unqualified points and the heat affected unqualified points are located are removed, the intervals that meet the processing requirements are selected, after removing all unqualified areas, the remaining continuous areas that meet all the set conditions are selected, the intervals represent that the workpiece part can be safely processed, and the processing available interval set is obtained.

[0072] S313: Call the processing available interval set, perform spatial positioning and geometric matching of the path segments in the processing path planning priority list, calculate the overlap degree and compatibility of the path segments and the available intervals, analyze the corresponding relationship between the path and the interval characteristics, identify the potential for processing parameter adjustment, map the adaptation relationship between the processing interval and the path, and use the formula:

[0073] ;

[0074] Calculate the evaluation index value of the processing adaptation relationship, and generate a processing adaptation relationship group;

[0075] wherein, represents the evaluation index value of the processing adaptation relationship, represents the spatial positioning parameter of the path segment and the available interval i, represents the spatial positioning parameter of the path segment and the available interval i, total number of representative path segments and available intervals, representative calculation adjustment constant;

[0076] The processing available interval set, for example, the set of multiple continuous regions that meet the processing requirements screened out, is called to perform spatial positioning and geometric matching of the path segments in the path planning priority list, each path in the path planning priority list is decomposed into multiple path segments, for example, a path is composed of 10 line segments, each line segment is regarded as a path segment, and then each path segment is compared and matched in space with all available intervals in the processing available interval set, for example, by calculating whether the endpoints of the path segment fall within the available interval and whether the overall shape of the path segment is compatible with the shape of the available interval, the overlap calculation and compatibility evaluation of the path segment and the available interval are performed, for each path segment and available interval, the overlap area and overlap rate, for example, the ratio of the overlap area to the path segment area, are calculated, and the geometric compatibility, for example, if the curvature of the path segment is too large and the available interval is linear, the compatibility is low, the corresponding relationship between the path and interval characteristics, for example, the relationship between the length, curvature of the path segment and the size, shape of the available interval, and the relationship between the defect type of the region where the path segment is located and the material uniformity, geometric consistency of the available interval, is analyzed, the processing parameter adjustment potential, for example, if a path segment is in a highly compatible available interval and the material uniformity of the available interval is high and the curvature change is gentle, the processing parameters such as laser power and scanning speed of the path segment have greater adjustment space, for example, the laser power can be appropriately increased to improve the processing efficiency, the adaptation relationship between the processing interval and the path is mapped, and the adaptation relationship between each path segment and the corresponding available interval is established according to the overlap, compatibility and parameter adjustment potential, using the formula: ; calculate the evaluation index value of the processing adaptation relationship;

[0077] wherein, represents the evaluation index value of the processing adaptation relationship, indicating the comprehensive degree of matching between the path segment and the available interval, the higher the value, the better the matching degree, represents the spatial positioning parameter of the path segment and the available interval i, for example, it can represent the center point coordinates of the path segment , or the average curvature of the path segment, represents the spatial positioning parameter of the path segment and the available interval i, for example, it can represent the center point coordinates of the available interval i , or the average curvature of the available interval i, here, and are characteristic vectors representing the path segment and the available interval, rather than a single numerical value, in order to simplify the calculation example, it is assumed that and respectively represent the eigenvalue of the path segment and the available interval in a certain dimension, for example, the average curvature value, or the X coordinate of the center point, in practical applications, and may be a multi-dimensional vector, for example, , wherein is the X coordinate of the center point of the path segment, is the average curvature of the path segment, is the corresponding value of the available interval, at this time the difference and square sum operations in the formula need to be extended to vector norm calculation, but for the sake of clarity of the example, single-dimensional features are still taken as an example, represents the total number of path segments and available intervals, in the calculation, represents the number of feature dimensions considered when evaluating the adaptation relationship between a certain path segment and an available interval, here indicates three feature dimensions of the X coordinate of the center point, the Y coordinate of the center point, and the average curvature, represents the calculation of the adjustment constant, which is used to adjust the sensitivity of the exponential term, and its value is determined by experiment or experience, for example, when and the difference is small, the exponential term contributes less to , when the difference is large, the contribution increases rapidly, experimental verification shows that when takes a value between 3 and 10, it can effectively distinguish between good and poor adaptation, here ;

[0078] The benefit of the formula is that by introducing the exponential term, it can punish those spatial positioning parameters with large differences, so that path segments and available intervals with high matching degree obtain higher evaluation values, and the formula considers the matching degree of spatial position and shape features, avoiding the one-sidedness of a single indicator, now an example is given, assuming that a path segment and an available interval need to be evaluated for their adaptation relationship, considering three feature dimensions, namely :

[0079] dimension 1: X coordinate ( ); dimension 2: Y coordinate ( ); dimension 3: average curvature ( );

[0080] Set the parameter values: the X coordinate of the center point of the path segment , the X coordinate of the center point of the available interval , the Y coordinate of the center point of the path segment , the Y coordinate of the center point of the available interval , the average curvature of the path segment , and the average curvature of the available interval , adjustment constant ;

[0081] Calculation process: for (X coordinate):

[0082] ; ; ;

[0083] Term 1 contribution ;

[0084] For (Y coordinate):

[0085] ; ; ;

[0086] Term 2 contribution ;

[0087] For (Average curvature): ; ; ;

[0088] Term 3 contribution ;

[0089] Total fitting index value ;

[0090] The results show that the fitting index value of the path segment and the available interval on the selected three characteristic dimensions is 0.278621, this value will be compared with the fitting index value of the path segment and the available interval, for quantifying the matching degree, the higher the fitting index value, the better the adaptability of the path segment in the available interval, this value will be used as an important basis for generating the machining adaptation relationship group, guiding the subsequent machining parameter collaborative control.

[0091] Please refer to Figure 5 , the generation step of the machining parameter collaborative control instruction set is specifically:

[0092] S411: According to the machining adaptation relationship group, extract the surface roughness requirement and topography precision index in the current machining interval, set the material removal rate target and thermal damage control threshold, perform quantitative evaluation of laser power adjustment range and response speed, detect the scanning galvanometer speed and acceleration performance, match the machining demand and response ability, and generate machining demand deviation data group;

[0093] The quantitative evaluation of the laser power adjustment range and the response speed refers to detecting and quantifying the dynamic adjustment ability of the laser power and the response speed of the laser to the adjustment instruction;

[0094] According to the machining adaptation relationship group, the surface roughness requirement and the topography accuracy index in the current machining interval are extracted, the surface roughness target value and the topography accuracy index corresponding to the interval are obtained from the machining feature distribution table according to the machining interval indicated by the machining adaptation relationship group, for example, the current interval is a high-precision machining area, the surface roughness requirement is Ra less than 0.8 microns, and the topography accuracy index is a maximum deviation less than 0.02 millimeters, the material removal rate target and the thermal damage control threshold are set, the material removal rate target in the laser machining process is set according to the material type, thickness and machining requirement, for example, for 2 millimeter thick 304 stainless steel, the target removal rate is set to 0.5 cubic millimeters per second, at the same time, according to the material thermal damage sensitivity and the demand to avoid micro-crack generation, the thermal damage control threshold is set, for example, the maximum temperature of the machining area is not allowed to exceed 800 degrees Celsius, and the heat affected zone width is not more than 0.3 millimeters, the laser power adjustment range and response speed quantitative evaluation are carried out, the laser power output range is quantified through actual test and calibration, for example, the laser power can be continuously adjusted between 100W and 1000W, and the response time required for the laser to change from receiving an adjustment instruction to the actual power output to stabilize is evaluated, for example, the response time for the power to change from 500W to 600W is 10 milliseconds, the scanning galvanometer speed and acceleration performance are detected, the maximum scanning speed of the scanning galvanometer is determined through experimental measurement, for example, the maximum scanning speed is 5000 millimeters per second, and the acceleration performance required from static acceleration to maximum speed or from maximum speed to static deceleration is determined, for example, the maximum acceleration is 200000 millimeters per second square, the set material removal rate target, thermal damage control threshold and laser power adjustment range, response speed, scanning galvanometer speed and acceleration performance are compared, for example, the current machining area requires high removal rate and strict thermal damage control, but the laser response speed is slow, indicating that the matching degree is low, for example, to achieve a removal rate of 0.5 cubic millimeters per second, theoretically 800W laser power is required, and the laser needs to be raised from the current power of 500W to 800W in 10 milliseconds, while the path segment only needs 5 milliseconds to complete, the response capability is insufficient, and the machining requirement deviation data group is generated.

[0095] S412: Call the machining requirement deviation data group, compare the parameter deviation values with the preset allowable range one by one, identify and classify the deviation items exceeding the allowable range, divide the matching degree between the machining requirement and the response capability into levels and represent the state, mark the key parameter matching insufficient area, and generate a matching state identification set;

[0096] The one-by-one comparison parameter deviation value and the preset allowable range refer to the numerical comparison of the laser power adjustment range deviation, response speed deviation, scanning galvanometer speed deviation and scanning galvanometer acceleration deviation included in the processing requirement deviation data set with the preset allowable range;

[0097] The processing requirement deviation data set is called to compare the laser power adjustment range deviation, response speed deviation, scanning galvanometer speed and acceleration deviation with the preset allowable range one by one, such as that the power deviation is 20W, the allowable range is ±50W, the response speed deviation is 5ms, the allowable range is ±2ms, the speed deviation is 50mm / s, the allowable range is ±100mm / s, and the acceleration deviation is 30000mm / s², the allowable range is ±50000mm / s². If the range is exceeded, such as the response speed deviation, it is marked as “insufficient response speed” and classified as a high-priority adjustment item; if the power deviation is within the range, it is not marked. According to the number and severity of the deviations, the matching level is divided: “high matching degree” if all are within the range, “medium matching degree” if a small amount is slightly exceeded, and “low matching degree” if it is significantly exceeded. For example, only the response speed is insufficient, it is marked as medium matching degree, and the specific insufficient item is recorded to generate a matching state identification set.

[0098] S413: According to the matching state identification set, verify the search of the processing interval number and the spatial position, perform the increment or decrement calculation of the required laser power of the current interval, set the acceleration or deceleration amplitude of the scanning speed, identify the power and speed adjustment instructions, and generate a processing parameter cooperative control instruction set;

[0099] Based on the matching state identification set, search for the processing interval spatial information, such as the starting point (100, 50) and the ending point (120, 55), locate the physical processing area, and calculate the required power increment or decrement, such as that the target removal rate is 0.5mm³ / s and requires 800W, and the current power is 500W, so the increment is +300W; if finer processing is required, the power can be reduced to 450W. Combined with the galvanometer performance and path geometry, the scanning speed adjustment amplitude is set, such as that the galvanometer allows an acceleration of 200000mm / s², the linear segment speed is 1000→1200mm / s, or the complex curve is 800→600mm / s. The power and speed adjustments are converted into digital instructions, such as “power +300W” and “speed +200mm / s”, and finally a processing parameter cooperative control instruction set is generated for real-time driving of power and galvanometer synchronous adjustment to ensure processing precision and efficiency.

[0100] Please refer to Figure 6 , and the generation steps of the optimal processing path and synchronous processing setting table are as follows:

[0101] S511: Call the processing parameter cooperative control instruction set, analyze the difference between the laser scanning head motion instruction and the real-time response time sequence, monitor the power adjustment and the scanning galvanometer cooperative action lag, measure the fluctuation range of the response time under the differentiated load conditions, use the formula:

[0102] ;

[0103] Calculate the path delay fluctuation characteristic value, generate the path delay fluctuation correlation coefficient;

[0104] Wherein, represents the path delay fluctuation characteristic value, represents the total number of sampling sequences, represents the time stamp of the laser scanning head motion instruction in the kth sampling sequence, represents the real-time response time stamp of the laser scanning head in the kth sampling sequence;

[0105] Call the processing parameter cooperative control instruction set, for example, a sequence containing instructions such as "laser power + 300W" and "scanning speed + 200mm / s", analyze the difference between the laser scanning head motion instruction and the real-time response time sequence, compare the motion instruction of the laser scanning head (i.e. scanning galvanometer) set in the instruction, for example, linear scanning from point A to point B at a speed of 200mm / s, with the actual feedback motion trajectory and time stamp of the laser scanning head, for example, the time instruction for the laser scanning head to move from point A to point B is 50ms, but the actual completion time is 53ms, there is a time sequence difference of 3ms, monitor the power adjustment and the scanning galvanometer cooperative action lag, monitor the lag between the time required for the actual power output to stabilize after the laser receives the power adjustment instruction (for example, "laser power + 300W") and the time when the scanning galvanometer starts to respond (for example, starts scanning), for example, after the power instruction is issued, the laser power stabilizes within 10ms, and the scanning galvanometer starts to move 2ms before the power stabilizes, so the lag is -2ms (indicating that the mirror is ahead of time), if the mirror moves 3ms after the power stabilizes, the lag is 3ms, measure the fluctuation range of the response time under the differentiated load conditions, repeatedly measure the response time of the laser scanning head under different processing load conditions (for example, processing workpieces of different thicknesses and different materials, or high-speed and low-speed scanning), and statistics the fluctuation range of the response time, for example, when processing 1mm thick stainless steel, the laser scanning response time fluctuates between 10ms and 12ms, and when processing 2mm thick stainless steel, the fluctuation range is between 12ms and 15ms, use the formula: ; Calculate the path delay fluctuation characteristic value, generate the path delay fluctuation correlation coefficient;

[0106] Wherein, representing the path delay fluctuation characteristic value, used to quantify the difference between the laser scanning head motion instruction and the real-time response timestamp under a certain sampling sequence, the larger the value, the greater the fluctuation, representing the total number of sampling sequences, indicating the number of samples for timing difference monitoring, here representing that 5 groups of data have been collected, which is the number of discrete samples for evaluating the difference between the laser scanning head motion instruction and the real-time response timestamp, representing the timestamp of the laser scanning head motion instruction in the kth sampling sequence, for example, the system time when the instruction is issued, representing the real-time response timestamp of the laser scanning head in the kth sampling sequence, for example, the feedback time when the laser head actually starts to execute the instruction;

[0107] The advantage of the formula is that by calculating the root mean square error of the instruction timestamp and the actual response timestamp, the stability and consistency of the laser scanning response in the processing process can be accurately reflected, high fluctuation characteristic value indicates that there are unstable factors, which need to be further optimized and coordinated control, low fluctuation characteristic value indicates that the system response is stable, which is conducive to ensuring the processing precision; now an example is given, assuming that 5 samples are taken, the following instruction timestamp and real-time response timestamp data are obtained:

[0108] Table 1: Path delay fluctuation sampling data table:

[0109]

[0110] Calculation process:

[0111] , ;

[0112] , ;

[0113] , ;

[0114] , ;

[0115] , ;

[0116] , ;

[0117] milliseconds;

[0118] The results show that under the current sampling sequence, the path delay fluctuation characteristic value of the laser scanning head motion instruction and the real-time response time stamp is 0.5099 milliseconds, which reflects the average deviation degree of the laser scanning system response, and is the key input for generating the path delay fluctuation correlation coefficient. If this value is higher than the preset threshold (for example, 0.4 milliseconds), it indicates that the response has a large fluctuation, and further system calibration or optimization is needed. The fluctuation characteristic value will be used to evaluate the processing stability of the path and determine whether it is suitable for synchronous processing.

[0119] S512: According to the path delay fluctuation correlation coefficient, analyze the complexity of the path geometry, and perform parameter mapping of the material thickness and hardness variation. Set the precision requirement and stability margin of the synchronous processing condition, evaluate the continuity and smoothness constraints between path segments, and select the path set that meets the synchronous processing condition to generate the synchronous processing judgment value interval.

[0120] According to the path delay fluctuation correlation coefficient, obtain the geometric feature parameters such as path curvature, number of turning points and curvature change rate, and divide them into "low", "medium" and "high" levels. Perform parameter mapping of the material thickness and hardness variation, provide the material thickness and hardness variation, map to the processing stability influence factor, set the precision requirement and stability margin of the synchronous processing condition, set the synchronous processing precision threshold according to the product precision level and process requirement, and set the stability margin to cope with fluctuations and external disturbances. Evaluate the continuity and smoothness constraints between path segments, evaluate the geometric variation or material property variation at the connection points of adjacent path segments, ensure the processing continuity and smoothness, select the path set that meets the synchronous processing condition, and filter out the paths that meet the synchronous processing condition. Generate the synchronous processing judgment value interval.

[0121] S513: Call the synchronous processing judgment value interval, perform response speed and precision matching of laser power adjustment, and perform acceleration limitation and smoothness comparison of scanning galvanometer speed variation. Optimize the transition time and seamless connection of path segment parameter switching, determine the optimal processing path and corresponding processing parameters, and output the optimal processing path and synchronous processing setting table.

[0122] The synchronous processing judgment value interval is called to filter out all paths meeting the synchronous processing conditions and their corresponding judgment value ranges, and the laser power adjustment response speed and accuracy are matched. In the synchronous processing judgment value interval, the power adjustment instruction in the processing parameter cooperative control instruction set is used for matching analysis with the evaluation of the laser response speed. For example, a certain path segment needs to be quickly switched from 500W power to 800W, and the laser response time is 10 milliseconds. The effective processing time of the path segment is only 8 milliseconds. The power adjustment response speed does not match the processing requirement and needs to be adjusted. The accuracy matching requires that the output power deviation after power adjustment is less than 1%. For example, the power deviation is kept within plus or minus 5W after adjustment, and the scanning galvanometer speed change acceleration limit and smoothness are compared. The scanning speed change requirement on the path segment, for example, from 1000mm / s to 1200mm / s, is compared with the evaluation of the maximum acceleration performance of the scanning galvanometer to ensure that the speed change does not exceed the physical limit of the galvanometer. For example, the required acceleration is 250000mm / s2, and the maximum acceleration of the galvanometer is 200000mm / s2, so it cannot be realized. The smoothness during the speed change process is evaluated to ensure that the acceleration change rate is within the allowed range to avoid vibration. The transition time optimization and seamless connection between path segments are performed for the parameter switching points between adjacent path segments in the same path. For example, the transition time during the switching process is optimized by preloading or early instruction method, and the power is smoothly transitioned from one value to the next value to avoid sudden changes. For example, the power is transitioned by setting a 0.1 second ramp-up / down time to ensure seamless connection during processing. For example, when the laser power and scanning speed reach the new target value, the next path segment starts to execute immediately to avoid processing defects at the junction of the path segments. The optimal processing path and corresponding processing parameters are determined. Based on the above matching and optimization results, the path with excellent performance in power response speed, accuracy, scanning galvanometer acceleration and transition between path segments is selected as the optimal processing path from the synchronous processing judgment value interval, and the corresponding accurate processing parameter combination is determined. For example, path 1 is selected as the optimal path, and the corresponding parameters are laser power 800W and scanning speed 1200mm / s. The optimal processing path and synchronous processing setting table are output.

[0123] The above is only a preferred embodiment of the present application, and does not limit the form of the present application. Any skilled person in the art can use the disclosed technical content to make changes or modifications to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the technical solution of the present application.

Claims

1. A method for planning combined paths and laser processing based on machine vision, characterized in that, Includes the following steps: S1: Real-time acquisition of images of the processing area through an industrial camera, extraction of the workpiece shape contour and measurement of geometric parameters, classification and spatial positioning of workpiece surface defects, identification of the physical and chemical properties of the workpiece material and analysis of its composition, and generation of a processing feature distribution table. S2: Based on the processing feature distribution table, set the start and end points of the multi-target processing path, optimize the connection order of multiple path segments, adjust the smoothing of path inflection points and the path in the obstacle avoidance area, and generate a path planning priority list. S3: Based on the path planning priority list, perform real-time measurement and trend capture of the workpiece surface curvature, and perform dynamic detection and diffusion assessment of the boundaries of the processing area and heat-affected area, dynamically adjust the range delineation and condition matching of the processing area, and establish a processing adaptation relationship group. The specific steps for generating the processing adaptation relationship group are as follows: S311: Call the path planning priority list to perform real-time scanning and three-dimensional reconstruction of the workpiece surface curvature, identify curvature gradients and abrupt change points, dynamically monitor the temperature field of the processing area and construct a heat diffusion model, perform real-time tracking and morphology evaluation of the boundary of the heat-affected area, summarize curvature change and heat-affected data, and obtain processing environment state parameters. S312: Based on the processing environment state parameters, set the upper and lower limits of the curvature variation range, define the allowable threshold and safety boundary of the heat-affected zone, compare and analyze the deviation of the regional parameter data point by point, identify and eliminate regions that do not meet the processing accuracy or material integrity requirements, screen the intervals that meet the processing requirements, and obtain the set of usable processing intervals. S313: Call the set of available processing intervals, perform spatial positioning and geometric matching of path segments in the priority list of processing path planning, calculate the overlap and compatibility of path segments and available intervals, analyze the correspondence between path and interval features, identify the potential for adjusting processing parameters, map the adaptation relationship between processing intervals and paths, calculate the evaluation index value of processing adaptation relationship, and generate processing adaptation relationship group. S4: Based on the processing adaptation relationship group, set the parameter thresholds for the current processing requirements and evaluate the execution capability, calibrate the delay time and stability prediction of the response capability, perform continuous stepless adjustment of laser power and synchronous adjustment of scanning speed, and output a set of collaborative control instructions for processing parameters. S5: Call the processing parameter collaborative control instruction set to perform real-time delay compensation and prediction of the processing path, analyze the fluctuation characteristics and error correction of the response time, perform dynamic correction of the path and correspondence of processing parameters, and output the optimal processing path and synchronous processing setting table. The optimal processing path and synchronous processing setting table includes path delay characteristics, synchronous processing time points, and response characteristics.

2. The method for combined path and laser processing planning based on machine vision according to claim 1, characterized in that, The processing feature distribution table includes workpiece number, shape feature identifier, defect distribution mark, and material property label; the path planning priority list includes path weight value, processing sequence identifier, and priority level; the processing adaptation relationship group includes curvature change range, heat-affected range identifier, and processing feasibility judgment value; and the processing parameter collaborative control instruction set includes power level code, speed adjustment direction, and cycle adjustment identifier.

3. The method for combined path and laser processing planning based on machine vision according to claim 1, characterized in that, The specific steps for generating the processing feature distribution table are as follows: S111: Real-time acquisition of images of the processing area via industrial cameras, input of images into a deep learning model for processing, extraction and mapping of pixel-level features, identification of workpiece contour boundaries, analysis of topology, identification of surface defect categories and spatial location, analysis of material spectral response characteristics and inference of physical properties, and generation of processing feature sequences. S112: Call the processing feature sequence to perform matching analysis between workpiece geometry and defect distribution pattern, identify risk defect areas and mark processing priorities, and generate a key processing area index set; S113: Based on the key processing area index set, identify geometrically consistent areas, verify material uniformity areas, mark non-critical defect areas, and perform processing stability assessment to generate a processing feature distribution table.

4. The method for combined path and laser processing planning based on machine vision according to claim 3, characterized in that, The specific steps for generating the path planning priority list are as follows: S211: Call the processing feature distribution table, extract the geometric shape features and boundary information of the key processing area, identify the defect type, size, and location information and quantify them, perform preliminary construction of multiple alternative paths and set key nodes, evaluate and convert the values ​​of path length, curvature, obstacle avoidance, and surface roughness requirements, and generate a path weight set. S212: Based on the path weight set, compare and analyze the weight values ​​of each path and determine their magnitudes, arrange the paths, determine the priority levels of the paths, and generate a path priority sequence. S213: Based on the path priority sequence, perform a conformity check between each path and the preset processing parameters, and identify and filter paths that do not meet the processing accuracy requirements, pose a risk of material damage, or have low processing efficiency, and generate a path planning priority list.

5. The method for combined path and laser processing planning based on machine vision according to claim 1, characterized in that, The specific steps for generating the collaborative control instruction set for the processing parameters are as follows: S411: Based on the processing adaptation relationship group, extract the surface roughness requirements and morphology accuracy indicators within the current processing range, set the material removal rate target and thermal damage control threshold, perform quantitative evaluation of the laser power adjustment range and response speed, detect the scanning galvanometer speed and acceleration performance, match processing requirements and response capabilities, and generate a processing requirement deviation data group. S412: Call the processing demand deviation data group, compare the parameter deviation values ​​with the preset allowable range one by one, identify and classify the deviation items that exceed the allowable range, classify and characterize the matching degree between processing demand and response capability in a hierarchical manner, mark the areas with insufficient matching degree of key parameters, and generate a matching status identifier set. S413: Based on the matching status identifier set, verify the retrieval and spatial location of the processing interval number, perform incremental or decremental calculations of the laser power required for the current interval, set the acceleration or deceleration of the scanning speed, identify power and speed adjustment commands, and generate a set of collaborative control commands for processing parameters.

6. The method for combined path and laser processing planning based on machine vision according to claim 5, characterized in that, The quantitative evaluation of the laser power adjustment range and response speed refers to detecting and quantifying the dynamic adjustment capability of the laser power and the response speed of the laser to the adjustment command. The step-by-step comparison of parameter deviation values ​​with the preset allowable range refers to comparing the laser power adjustment range deviation, response speed deviation, scanning galvanometer speed deviation, and scanning galvanometer acceleration deviation included in the processing requirement deviation data group with the preset allowable range values ​​respectively.

7. The method for combined path and laser processing planning based on machine vision according to claim 1, characterized in that, The specific steps for generating the optimal processing path and synchronous processing setting table are as follows: S511: Call the processing parameter coordinated control instruction set, analyze the difference between the laser scanning head motion command and the real-time response timing, monitor the power adjustment and scanning galvanometer coordinated action lag, measure the fluctuation range of the response time under different load conditions, calculate the path delay fluctuation characteristic value, and generate the path delay fluctuation correlation coefficient. S512: Based on the path delay fluctuation correlation coefficient, analyze the complexity of the path geometry, perform parameter mapping of material thickness and hardness changes, set the accuracy requirements and stability margin of synchronous processing conditions, evaluate the continuity and smoothness constraints between path segments, screen the path set that meets the synchronous processing conditions, and generate the synchronous processing judgment value range. S513: Call the synchronous processing judgment value range, match the response speed and accuracy of laser power adjustment, and compare the acceleration limit and stability of the scanning galvanometer speed change. Optimize the transition time and seamlessly connect the parameter switching between path segments, determine the optimal processing path and corresponding processing parameters, and output the optimal processing path and synchronous processing setting table.

Images