Laser cutting path optimization method of laser cutting equipment

Abstract

The invention discloses a laser cutting path optimization method of laser cutting equipment, and relates to the technical field of cutting path optimization. According to the laser cutting path optimization method of the laser cutting equipment, geometric primitive data of a to-be-cut workpiece is obtained, contour analysis processing is carried out, and a contour geometric data set of the to-be-cut workpiece is extracted; cutting sequence planning processing is conducted on the contour geometric data set of the to-be-cut workpiece based on a genetic algorithm, and an initial cutting path of the to-be-cut workpiece is generated; multi-target collaborative optimization processing is conducted on the initial cutting path of the to-be-cut workpiece, and the optimized cutting path of the to-be-cut workpiece is obtained.According to the method, the optimized cutting path of the to-be-cut workpiece is verified, the optimized verification cutting path of the to-be-cut workpiece is generated, and therefore the machining efficiency is remarkably improved while the machining precision is guaranteed; therefore, the reliability of the cutting process is comprehensively improved.

Description

Technical Field

[0001] This invention relates to the field of cutting path optimization technology, specifically a laser cutting path optimization method for laser cutting equipment. Background Technology

[0002] In the industrial manufacturing field, laser cutting is widely used in various industries such as sheet metal processing, aerospace parts manufacturing, and automotive parts production due to its advantages such as high cutting precision, high speed, and small heat-affected zone. As the manufacturing industry upgrades towards high precision, high efficiency, and low cost, the contours of the workpieces to be cut are becoming increasingly complex, which places higher demands on the planning and optimization of laser cutting paths. The rationality of the cutting path directly determines the cutting efficiency, workpiece processing accuracy, equipment energy consumption, and the degree of workpiece deformation. Therefore, laser cutting path optimization has become an optimization direction to improve the overall performance of laser cutting processing.

[0003] In existing technologies, laser cutting path optimization methods have gradually evolved from early manual planning to algorithm-driven automatic planning. Currently, the mainstream optimization approaches are mostly based on intelligent optimization algorithms such as genetic algorithms and simulated annealing algorithms. By solving path sorting problems similar to the traveling salesman problem, the cutting sequence is optimized to shorten idle travel time and improve processing efficiency.

[0004] The limitations of existing technologies include at least the following issues: Firstly, existing technologies are prone to node redundancy or excessive fitting deviation during contour loop extraction, leading to errors in the geometric datum for subsequent path planning. Secondly, they neglect key factors such as the cutting sequence constraints of nested contours and workpiece deformation caused by path heat accumulation during cutting, easily causing workpiece stress concentration deformation and slag accumulation. Thirdly, the optimization objectives are one-sided, making it difficult to accurately balance mutually restrictive objectives such as efficiency improvement, accuracy assurance, and workpiece stability. Fourthly, the optimized path lacks a verification mechanism, failing to verify the contour closure accuracy and idle stroke interference of the path, leading to feasibility issues such as collisions between the cutting head and workpiece chips, and poor workpiece surface quality, thus reducing the reliability of cutting. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a laser cutting path optimization method for laser cutting equipment, which solves the problems of large geometric reference error, neglect of process constraints, difficulty in balancing multiple objectives, and low reliability caused by lack of verification in existing technologies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a laser cutting path optimization method for a laser cutting device, comprising the following steps: acquiring geometric data of the workpiece to be cut and performing contour parsing processing to extract the contour geometric dataset of the workpiece to be cut; performing cutting sequence planning processing on the contour geometric dataset of the workpiece to be cut based on a genetic algorithm to generate an initial cutting path for the workpiece to be cut; performing multi-objective collaborative optimization processing on the initial cutting path of the workpiece to be cut to obtain an optimized cutting path for the workpiece to be cut; and performing verification processing on the optimized cutting path of the workpiece to be cut to generate an optimized verification cutting path for the workpiece to be cut.

[0007] Further, the specific steps for extracting the contour geometry dataset of the workpiece to be cut are as follows: read the geometric data of the workpiece to be cut and perform recognition processing to obtain several contour geometry rings of the workpiece to be cut; perform spatial inclusion processing and geometric feature extraction processing on each contour geometry ring of the workpiece to be cut to obtain the ring ID, nesting level and geometric feature parameter set of each contour geometry ring of the workpiece to be cut.

[0008] Further, the specific steps to obtain several contour geometric loops of the workpiece to be cut are as follows: read the geometric data of the workpiece to be cut and preprocess it; construct the primitive connection graph of the workpiece to be cut based on the preprocessed geometric data of the workpiece to be cut; perform loop detection processing on the primitive connection graph of the workpiece to be cut based on the depth-first search method to generate several contour geometric loops of the workpiece to be cut.

[0009] Further, the specific steps for generating the initial cutting path of the workpiece to be cut are as follows: Based on the ring ID of each contour geometric ring of the workpiece to be cut, an initial path population of the workpiece to be cut is generated, including several candidate initial paths, and each candidate initial path is formed by randomly associating and splicing each contour geometric ring; Based on the geometric feature parameter set of each contour geometric ring of each candidate initial path of the workpiece to be cut, a fitness model of each candidate initial path of the workpiece to be cut is constructed, and the path fitness score of the corresponding candidate initial path is extracted; Based on the path fitness score of each candidate initial path of the workpiece to be cut, the initial path population of the workpiece to be cut is iteratively processed until the preset convergence condition is met, and the initial cutting path of the workpiece to be cut is generated.

[0010] Furthermore, the specific steps for constructing the fitness model for each candidate initial path of the workpiece to be cut are as follows: Based on the geometric feature parameter set of each contour geometric loop of each candidate initial path of the workpiece to be cut, extract the path evaluation dataset of the corresponding candidate initial path, including empty travel value and nested conflict number value; Based on the path evaluation dataset of each candidate initial path of the workpiece to be cut, construct the fitness model of the corresponding candidate initial path.

[0011] Furthermore, the specific steps to obtain the optimized cutting path of the workpiece to be cut are as follows: Based on the initial cutting path and contour geometry dataset of the workpiece to be cut, construct a path constraint optimization model for the workpiece to be cut; perform dynamic constraint iterative optimization on the path constraint optimization model for the workpiece to be cut to obtain the Pareto elite solution set of the workpiece to be cut; and filter the Pareto elite solution set of the workpiece to be cut based on preset decision rules to generate the optimized cutting path of the workpiece to be cut.

[0012] Furthermore, the specific steps for constructing the path constraint optimization model of the workpiece to be cut are as follows: Based on the initial cutting path and contour geometry dataset of the workpiece to be cut, the total cutting time and path heat accumulation value of the workpiece to be cut are extracted respectively; the total cutting time and path heat accumulation value of the workpiece to be cut are defined as the main optimization objective and the collaborative constraint objective of the workpiece to be cut respectively; the main optimization objective and the collaborative constraint objective of the workpiece to be cut are integrated to construct the path constraint optimization model of the workpiece to be cut.

[0013] Furthermore, the specific steps for obtaining the Pareto elite solution set of the workpiece to be cut are as follows: mark the initial cutting path of the workpiece to be cut as the current cutting path, and set the constraint range and the Pareto empty set; perform inner-layer local iterative search processing on the current cutting path based on the set constraint range and the path constraint optimization model of the workpiece to be cut to update the current cutting path, and update the Pareto empty set synchronously; when the set constraint range reaches the preset termination condition, mark the updated Pareto empty set as the Pareto elite solution set of the workpiece to be cut.

[0014] Furthermore, the specific steps for generating the optimized verification cutting path for the workpiece to be cut are as follows: perform multi-target verification processing on the optimized cutting path of the workpiece to be cut; determine whether the optimized cutting path of the workpiece to be cut meets the preset verification qualification conditions; if the optimized cutting path of the workpiece to be cut meets the preset verification qualification conditions, then mark the optimized cutting path of the workpiece to be cut as the optimized verification cutting path; if the optimized cutting path of the workpiece to be cut does not meet the preset verification qualification conditions, then perform correction processing on the optimized cutting path of the workpiece to be cut to generate the optimized verification cutting path of the workpiece to be cut.

[0015] Furthermore, the specific steps for multi-target verification processing are as follows: Based on the contour geometry dataset of the workpiece to be cut, perform closure analysis on the optimized cutting path of the workpiece to be cut, and extract the contour closure accuracy set of the workpiece to be cut; Based on the contour geometry dataset of the workpiece to be cut, perform ray interferometry detection processing on the optimized cutting path of the workpiece to be cut, and extract the interference state identifier set of the workpiece to be cut.

[0016] The present invention has the following beneficial effects:

[0017] (1) The laser cutting path optimization method of this laser cutting equipment significantly improves the accuracy of the cutting path through the whole process design of structured contour analysis, process constraint adaptation, multi-objective collaborative optimization and closed-loop verification. By preprocessing geometric data and constructing primitive connection graph, a non-redundant contour geometric loop can be accurately generated and a high-precision contour geometric dataset can be constructed to provide a solid geometric benchmark for path planning. In the cutting sequence planning, a fitness model is constructed to ensure that the path strictly follows the process constraint that the sub-loop precedes the parent loop, avoiding stress concentration. Based on the path constraint optimization model, the Pareto elite solution set is generated by dynamic constraint iterative optimization to achieve a precise balance between efficiency and deformation control. After multi-objective verification and correction, the path is free from interference and the closure accuracy meets the standard, thus comprehensively improving the reliability of the cutting process.

[0018] (2) The laser cutting path optimization method of this laser cutting equipment, through intelligent algorithm optimization and precise adaptation of process parameters, significantly improves processing efficiency while ensuring processing accuracy. Based on the genetic algorithm, an initial path population containing random and heuristic strategies is generated. Combined with the fitness model iterative screening, the optimal initial cutting path can be obtained quickly, shortening the path planning cycle. During the dynamic constraint iterative optimization process, the total cutting time is precisely optimized through the inner layer local search. At the same time, the path heat accumulation value is strictly controlled to reduce slag accumulation and workpiece deformation, improve the cutting surface quality and dimensional accuracy. The Pareto elite solution set combined with the preset decision rules can adapt to the needs of different processing scenarios and output customized optimal paths. The closed-loop verification and correction mechanism ensures that the optimized path can be directly used for actual processing, reducing the risk of failures such as cutting head collision and chip interference.

[0019] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0020] Figure 1 This is a flowchart of a laser cutting path optimization method for a laser cutting device according to the present invention.

[0021] Figure 2 This is a flowchart illustrating the specific steps involved in obtaining the optimized cutting path for the workpiece to be cut in the laser cutting path optimization method of a laser cutting device according to the present invention. Detailed Implementation

[0022] Please see Figure 1This invention provides a technical solution: a laser cutting path optimization method for a laser cutting device, comprising the following steps: acquiring geometric data of the workpiece to be cut and performing contour parsing processing to extract the contour geometric dataset of the workpiece to be cut; performing cutting sequence planning processing on the contour geometric dataset of the workpiece to be cut based on a genetic algorithm to generate an initial cutting path for the workpiece to be cut; performing multi-objective collaborative optimization processing on the initial cutting path of the workpiece to be cut to obtain an optimized cutting path for the workpiece to be cut; and performing verification processing on the optimized cutting path of the workpiece to be cut to generate an optimized verification cutting path for the workpiece to be cut.

[0023] Specifically, the geometric data metadata is effective data used to accurately define the outer contour shape and internal hole features of the workpiece to be cut. Its types include, but are not limited to, straight lines, arcs, elliptical arcs, and third-order Bézier curves. It specifically includes the core definition parameters of each type of primitive, including but not limited to the start / end coordinates of straight lines, the center / radius / start / end angle of arcs, the major / minor axis / center / rotation angle / start / end angle of elliptical arcs, and the control point coordinate array of third-order Bézier curves. The specific steps for extracting the contour geometry dataset of the workpiece to be cut are as follows:

[0024] Read the geometric metadata of the workpiece to be cut and perform recognition processing to obtain several contour geometric loops of the workpiece to be cut; perform spatial inclusion processing and geometric feature extraction processing on each contour geometric loop of the workpiece to be cut to obtain the loop ID, nesting level and geometric feature parameter set of each contour geometric loop of the workpiece to be cut, specifically as follows:

[0025] Each contour geometric loop is assigned a unique identifier (denoted as loop ID), and an inclusion tolerance is set, which is consistent with the geometric connection tolerance (i.e., 0.005mm) to avoid logical conflicts caused by multiple criteria. A ray casting method is then used to perform pairwise inclusion relationship analysis on all contour geometric loops.

[0026] Traverse all contour geometric rings, arbitrarily select two non-repeating contour geometric rings as objects to be judged, denoted as ring A and ring B. Randomly select a discrete node from the node coordinate sequence of ring A as the ray emission point, and emit an infinitely long ray in the preset positive X-axis direction. Construct its boundary contour based on the node coordinate sequence of ring B, and count the number of intersections between the emitted ray and the boundary contour of ring B. At the same time, calculate the minimum Euclidean distance between all discrete nodes of ring A and the boundary contour of ring B. If the number of intersections between the ray and the boundary of ring B is odd, and the minimum Euclidean distance between all discrete nodes of ring A and the boundary of ring B is ≥ the inclusion judgment tolerance (ensuring that the two rings do not cross or overlap), then it is determined that "ring A is completely nested and contained by ring B"; otherwise, if the number of intersections is even or the minimum distance is < the inclusion judgment tolerance, then it is determined that "the two rings have no inclusion relationship".

[0027] After completing pairwise traversal of all contour geometric rings according to the above rules, establish a bidirectional nested association identifier for each contour geometric ring: record the direct parent ring ID of each ring: if a ring is contained by multiple rings, only record the ID of the outermost ring that directly contains it (i.e., the containing ring IDs that differ by only 1 in level). If there is no directly containing ring, the direct parent ring ID is set to empty; record the list of direct child ring IDs of each ring: if a ring directly contains multiple rings, organize the IDs of these contained rings into a list and store it. If there is no directly contained ring, the list is empty.

[0028] Based on the bidirectional nested association identifier, the nesting level of each ring is determined according to the rule of "parent ring level + 1": the contour geometry ring with an empty direct parent ring ID is the outermost nested ring (level marked L0), corresponding to the outer contour of the workpiece to be cut; the contour geometry ring with the L0 level ring ID as its direct parent ring ID is the first-level nested ring (level marked L1), corresponding to the first-level inner hole of the workpiece to be cut; the contour geometry ring with the L1 level ring ID as its direct parent ring ID is the second-level nested ring (level marked L2), corresponding to the second-level inner hole of the workpiece to be cut; and so on, until all valid contour geometry rings are marked with a unique nesting level.

[0029] For each geometric ring contour, based on its existing node coordinate sequence, a set of geometric feature parameters is extracted, namely: Centroid coordinates: calculated using the polygon centroid calculation formula (i.e., weighted average method) based on the node coordinate sequence of the geometric ring contour; Minimum circumscribed rectangle boundary coordinates: by traversing all discrete node coordinates of the ring, extracting the minimum, maximum, minimum, and maximum values ​​of the X-coordinate, Y-coordinate, and constructing the boundary parameters of the minimum circumscribed rectangle; Total contour path length: by traversing the node coordinate sequence of the ring, calculating the length of the line segment between two adjacent discrete nodes according to the Euclidean distance formula, and summing all line segment lengths to obtain the total contour path length; Enclosed area: calculated using the shoelace formula based on the node coordinate sequence of the ring; Maximum contour curvature: by traversing the node coordinate sequence of the ring, calculating the curvature at each discrete node (derived based on the angle between the node and the two line segments formed by the two adjacent nodes and the length of the line segments), and extracting the maximum value among all curvature values ​​as the maximum contour curvature.

[0030] Each contour geometric ring is associated and bound one by one with its ring ID, bidirectional nested association identifier (direct parent ring ID / direct child ring ID list), nesting level, geometric feature parameter set, node coordinate sequence, and primitive sequence to form a contour geometric dataset.

[0031] The specific steps to obtain several contour geometric rings of the workpiece to be cut are as follows: Read the geometric data of the workpiece to be cut and perform preprocessing, that is, for each type of graphic element in the target graphic element set, extract its discrete node coordinate sequence according to the preset sampling density. For example: for straight line graphic elements, the discrete nodes directly take its start point and end point (no additional sampling is required to avoid redundancy); for circular arc / elliptical arc graphic elements, according to its start and end angles, sample discrete points at equal angular intervals (in this embodiment, the interval is preferably 1°, and the sampling density can be adjusted according to the workpiece accuracy requirements to ensure that the distance between adjacent discrete points is ≤0.01mm); for third-order Bézier curves: based on its parametric equation (parameter t∈[0,1]), sample discrete points at equal parameter intervals (in this embodiment, the t interval is preferably 0.01 to ensure that the curve shape fitting error is ≤0.005mm); and include the discrete sampling nodes of each graphic element, together with its own start point and end point, into the key node set and record the original graphic element identifier to which each node belongs.

[0032] Based on the preprocessed geometric data of the workpiece to be cut, an undirected primitive connection graph of the workpiece is constructed. That is, a geometric connection tolerance (e.g., 0.005mm) is set, the key node set is traversed, and the Euclidean distance between any two nodes of different primitives is calculated. If the Euclidean distance is less than the geometric connection tolerance, it is determined that the two primitives form a geometric connection at this node. Based on all the determined connection relationships, an undirected primitive connection graph is constructed, that is, with a single primitive as a vertex and the geometric connection relationship between primitives as an edge, and the corresponding association information between vertices (primitives) and edges (connections) is stored.

[0033] Based on the depth-first search method, loop detection is performed on the (undirected) primitive connection graph of the workpiece to be cut, generating several contour geometric loops of the workpiece. Specifically: all primitives in the primitive connection graph are initialized; any unvisited primitive is selected as the starting vertex; traversal begins along the edges of the primitive connection graph, simultaneously recording the traversal path (including primitive sequence and node coordinate sequence); when the next vertex (primitive) to be visited on the traversal path satisfies the tolerance connection condition (i.e., the Euclidean distance between the two vertices < geometric connection tolerance) with a visited non-direct predecessor vertex in the path, it is marked as a potential closed contour loop; calculation... The Euclidean distance between the beginning and end nodes of the path (the starting and ending nodes after connecting all primitives) is calculated. If this distance is less than the geometric connection tolerance, a valid contour geometric loop is confirmed. All primitives contained in the valid contour geometric loop are removed from the "unvisited primitive set". The primitive sequence, node coordinate sequence and closure mark are stored in traversal order to form an independent contour geometric loop. The above steps are repeated until all primitives in the primitive connection graph are empty. For isolated primitives not contained in any closed loop, they are determined to be invalid data (not required for contour construction) and such primitives are automatically removed to obtain several contour geometric loops.

[0034] In this implementation scheme, differentiated preprocessing rules are formulated for different primitives such as straight lines, arcs, and Bézier curves. This avoids node redundancy and ensures that the curve fitting error is ≤0.005mm, accurately restoring the workpiece contour shape. By using primitive connection graph construction and depth-first search, isolated invalid primitives can be automatically eliminated, and complete and valid contour geometric loops can be efficiently extracted. The ray method with uniform tolerance is used to determine the nesting relationship. Combined with bidirectional association identifiers and hierarchical markers, the correspondence between parent and child loops is clarified to avoid logical conflicts. At the same time, multi-dimensional feature parameters such as centroid coordinates and maximum curvature are extracted to form a structured contour geometric dataset. This provides a precise geometric benchmark for cutting sequence planning and multi-objective optimization, ensuring the accuracy and stability of subsequent processing.

[0035] Specifically, the steps for generating the initial cutting path of the workpiece to be cut are as follows: Based on the ring ID of each contour geometric ring of the workpiece to be cut, an initial path population is generated, including several candidate initial paths. Each candidate initial path is formed by randomly associating and splicing the contour geometric rings. Specifically:

[0036] The initial population size is set to N, which can be configured according to the number of contours to be cut and the computing resources. The value can be 100. Then, from the completed contour geometry dataset, the unique identity of all contour geometric rings is extracted and denoted as the ring ID set. This set constitutes the gene pool for individual encoding in the genetic algorithm. Each individual (i.e. a candidate cutting path scheme) will be directly represented by a permutation sequence of ring IDs.

[0037] Complete random generation is performed. This part is used to generate 80% of the individuals in the population. For each individual to be generated, the operation process is as follows: The complete set of ring IDs is regarded as a list. The classic Fisher-Yates shuffle algorithm is used to uniformly and randomly shuffle the list. This algorithm traverses the list and randomly selects an element equal to or behind the current position for each position and swaps them, thereby ensuring that the probability of generating each possible permutation sequence is equal. By performing this random shuffling operation multiple times independently, a batch of random cut sequences uniformly distributed in the solution space can be obtained, which effectively guarantees the diversity of the initial population genes.

[0038] Heuristic rule generation is performed to generate individuals that make up 20% of the population. This aims to inject solutions with a certain priority quality and guide the optimization direction. In this implementation, the "nearest neighbor greedy method" is used as the heuristic rule. The steps to generate such an individual are as follows: Randomly select an ID from the ring ID set as the starting cutting contour; query the contour geometry dataset to obtain the centroid coordinates of the current cutting contour; calculate the Euclidean distance between the centroid of each corresponding contour and the current centroid in the remaining ring ID set; select the ring ID with the smallest distance as the next cutting target; in the fourth step, add the ID to the sequence and set it as the new current cutting contour. Repeat this process until all ring IDs are sorted according to this nearest neighbor rule. By changing different random starting points and repeating the above process, multiple cutting sequences constructed based on the local nearest neighbor principle can be generated. These sequences usually have relatively short potential empty journeys, providing a high level of initial fitness starting point for the population.

[0039] The N ring ID permutations generated by the above hybrid strategy are normalized and encapsulated to form an initial population data structure that can be directly manipulated by the genetic algorithm. This population is represented in the program as a list containing N elements, each element of the list corresponding to an individual. This individual is a structured data object that encapsulates at least the following information: a unique identifier for the individual, used to distinguish different individuals in the population; and a gene sequence, which is an array that stores all ring IDs in sequence, fully defining the cutting path scheme represented by the individual.

[0040] Based on the geometric feature parameter set of each contour geometric loop of each candidate initial path of the workpiece to be cut, a fitness model of each candidate initial path of the workpiece to be cut is constructed, and the path fitness score of the corresponding candidate initial path is extracted; based on the path fitness score of each candidate initial path of the workpiece to be cut, the initial path population of the workpiece to be cut is iteratively processed until a preset convergence condition is met, and the initial cutting path of the workpiece to be cut is generated, specifically as follows:

[0041] Invoke the preset genetic algorithm iteration parameters (the parameters can be adapted according to the contour complexity of the workpiece to be cut, which is preferred in this embodiment):

[0042] Maximum number of iterations: 100~200 (200 if the number of contour geometry loops is greater than 20, otherwise 100);

[0043] Crossover probability: 0.7~0.9 (preferably 0.85, to balance population diversity and iteration efficiency).

[0044] Mutation probability: 0.01~0.05 (preferably 0.03 to avoid excessive mutations that could destroy high-quality paths);

[0045] Elite retention rate: 10% (retaining the top 10% of candidate paths with fitness scores in each generation to avoid losing the optimal solution);

[0046] Convergence threshold: 0.001 (convergence is defined as the improvement in optimal fitness over multiple consecutive generations being less than this value).

[0047] Based on the fitness scores of each candidate initial path in the current initial path population, the parent population used for subsequent crossover and mutation is selected. The specific process is as follows:

[0048] Calculate the sum of fitness scores for all candidate paths in the current population; calculate the selection probability of each candidate path (i.e., the ratio of the fitness score of each candidate path to the sum of fitness scores; the higher the fitness, the greater the selection probability); use roulette wheel random sampling based on the selection probability to select candidate paths of equal size to the current population to form the parent population; extract the top 10% of candidate paths in the current population based on their fitness scores and retain them directly as "elite individuals" to the offspring population without participating in sampling;

[0049] Crossover operation (sequential crossover method, adapted to the permutation characteristics of the cutting sequence)

[0050] The candidate paths in the parent population are cross-crossed to generate offspring paths. Specifically, the parent population is randomly paired to obtain several parent pairs (denoted as parent 1 and parent 2). For each parent pair, two positions in the path cutting sequence are randomly selected as cross-cross intervals (such as the positions of the 2nd to 5th ring IDs). The ring ID sequence in the cross-cross interval of parent 1 is completely copied to the corresponding interval of the offspring path. Starting from the starting position of the path cutting sequence of parent 2, the ring IDs that do not appear in the cross-cross interval of the offspring are extracted in sequence and filled into the remaining interval of the offspring path.

[0051] For the completed child path, check whether its cutting order conforms to the nesting constraint of "child loop before direct parent loop" (call the bidirectional nested association identifier in the contour geometry dataset): if it conforms to the constraint, the child path is retained; if it violates the constraint, the intersection interval is reselected, and the steps of selection operation, intersection operation, and mutation operation are repeated until a compliant child path is generated.

[0052] Mutation operation (random exchange mutation method to avoid local optima): Mutation operation is performed on the offspring paths generated by the crossover operation. That is, each offspring path is traversed, and the mutation probability \(P_m\) is randomly determined to perform mutation on it. For the offspring path to be mutated, two different gene positions (i.e., the positions of two ring IDs) in the path cutting sequence are randomly selected. The ring IDs corresponding to these two gene positions are swapped to obtain the mutated offspring path. The mutated offspring path is checked to see if it meets the nesting constraints: if it meets the constraints, the mutation result is retained; if it violates the constraints, the swap is canceled, gene positions are reselected and mutation is performed until a compliant offspring path is generated.

[0053] Population Update and Convergence Determination: Merge "elite individuals" with compliant offspring paths generated after crossover and mutation to form a new generation of initial path population; call the fitness model constructed above to recalculate the fitness scores of all candidate paths in the new generation population; record the optimal fitness score of the current generation and compare it with the optimal fitness scores of the previous 20 generations: if the current iteration count reaches the maximum, the iteration is terminated; if the improvement of the optimal fitness score of the previous 20 generations (i.e., the absolute value of the difference between the fitness score of the 20th generation and the fitness score of the first generation is less than a preset threshold), the iteration is terminated; if the termination / convergence conditions are not met, repeat the selection, crossover, and mutation operations until the conditions are met.

[0054] After the iteration terminates, the path with the highest fitness score is selected from all candidate initial paths recorded during the iteration process and used as the initial cutting path for the workpiece to be cut. At the same time, the path cutting sequence, node coordinate sequence of each contour geometric ring, temporary empty travel path information, etc. corresponding to this path are stored together.

[0055] The specific steps for constructing the fitness model for each candidate initial path of the workpiece to be cut are as follows: Based on the geometric feature parameter set of each contour geometric ring of each candidate initial path of the workpiece to be cut, extract the path evaluation dataset of the corresponding candidate initial path, including the empty travel value and the nested conflict number value. Specifically, it is as follows: traverse the path cutting sequence of the candidate initial path, and call the centroid coordinates of each contour geometric ring in the geometric feature parameter set in the order of the two adjacent contour geometric rings in the sequence; use the Euclidean distance formula to calculate the distance between the centroid coordinates of two adjacent contour geometric rings as the single empty travel distance between the adjacent rings; accumulate the single empty travel distances of all adjacent rings to obtain the empty travel value of the candidate initial path.

[0056] Nested conflict count extraction: The nesting association identifiers (direct parent ring ID / direct child ring ID list) and nesting level information of each contour geometric ring in the contour geometry dataset are called, and the path cutting sequence of the candidate initial path is traversed. If there is a case in the sequence where "the cutting order of the direct parent ring is before its corresponding child ring", it is counted as 1 nested conflict. The number of nested conflicts in the sequence is accumulated to obtain the nested conflict count of the candidate initial path.

[0057] Based on the path evaluation dataset of each candidate initial path of the workpiece to be cut, a fitness model for the corresponding candidate initial path is constructed. Specifically, the empty travel value and nested conflict number value of each candidate initial path of the workpiece to be cut are normalized (their values ​​are mapped to between 0 and 1), and a fitness model for the corresponding candidate initial path is constructed accordingly.

[0058] The fitness model for a certain candidate initial path is as follows: ;in, The path fitness score is a candidate initial path for the workpiece to be cut. The empty travel value of a candidate initial path for the workpiece to be cut, after normalization. This refers to the empty travel adjustment coefficient stored in the database (its value range can be 0.5~0.8). This represents the nested conflict count of a candidate initial path for a workpiece to be cut, after normalization. This refers to the nested conflict adjustment coefficient stored in the database (its value can range from 0.2 to 0.5). .

[0059] In this implementation scheme, a hybrid strategy of "80% random generation + 20% heuristic generation" is adopted. This strategy relies on the Fisher-Yates shuffle algorithm to ensure the genetic diversity of the population, while the nearest neighbor greedy method is used to inject high-quality solutions for low-altitude routes, balancing exploration and utilization efficiency. During the genetic algorithm iteration, crossover and mutation operations adapted to the characteristics of the cutting sequence are combined with a 10% elite retention mechanism to avoid the loss of the optimal solution. At the same time, nested constraints are checked throughout the process to ensure that the path conforms to the rule of "child cycle before parent cycle". The fitness model integrates two indicators, empty routes and nested conflicts, to accurately evaluate the quality of the path. The final generated initial path has both efficiency potential and process compliance, which greatly improves the convergence speed of subsequent optimization.

[0060] Specifically, such as Figure 2 As shown, the specific steps to obtain the optimized cutting path of the workpiece to be cut are as follows: Based on the initial cutting path and contour geometry dataset of the workpiece to be cut, construct the path constraint optimization model of the workpiece to be cut; perform dynamic constraint iterative optimization on the path constraint optimization model of the workpiece to be cut to obtain the Pareto elite solution set of the workpiece to be cut.

[0061] Based on preset decision rules, the Pareto elite solution set of the workpiece to be cut is filtered to generate an optimized cutting path for the workpiece. Specifically, the preset decision rules are as follows: Calculate the Euclidean distance of each path in the solution set to a virtual "ideal point". The coordinates of the ideal point are composed of the minimum total cutting time and the minimum path heat accumulation value in the solution set. Select the path closest to the ideal point as the final selection. Traverse and calculate all paths in the Pareto elite solution set. Based on the calculation results of the selected rules, determine an optimal path and use it as the optimized cutting path for the workpiece. Output all relevant information of the path, including but not limited to: the complete contour cutting sequence, the total cutting time and path heat accumulation value corresponding to the sequence, and the cutting trajectory point sequence of each contour.

[0062] The specific steps for constructing the path constraint optimization model for the workpiece to be cut are as follows: Based on the initial cutting path and contour geometry dataset of the workpiece to be cut, the total cutting time and path heat accumulation value of the workpiece to be cut are extracted respectively, specifically as follows:

[0063] Based on the initial cutting path of the workpiece to be cut, the path cutting sequence is determined. For each pair of adjacent contour geometric rings in the path cutting sequence, the geometric feature parameter sets of the two contours are extracted from the previously constructed contour geometric dataset, and their centroid coordinates are obtained from them. The Euclidean distance between the two centroid coordinates is calculated, and this distance is used as the approximate travel length of the empty movement between the two contours.

[0064] Divide the approximate travel length calculated above by a preset laser cutting equipment idle speed (this speed is determined by the equipment performance parameters and is usually a fixed value) to obtain the single-segment idle travel time between the two adjacent contours; calculate the single-segment idle travel time according to the path cutting sequence, and finally sum all the single-segment idle travel times to obtain the total idle travel time of the entire cutting path.

[0065] For each contour geometric loop in the path cutting sequence, the total path length of the contour geometric loop is extracted from the corresponding geometric feature parameter set of the contour geometric dataset. A suitable cutting process speed is determined for each contour (a process parameter database can be preset, the core of which is a lookup table whose index key is a combination of material type and plate thickness. Each key corresponds to a set of basic process parameters, which includes a recommended basic cutting speed. When processing a workpiece to be cut, the system first queries the database based on the overall material type and plate thickness of the workpiece and determines the basic cutting speed applicable to all contours). The total path length of each contour is divided by its matching cutting process speed to obtain the time required to cut the single contour. The individual cutting times of all contours in the path cutting sequence are summed, and the sum is the total effective cutting time of the entire cutting path. The calculated total idle travel time is added to the total effective cutting time, and the sum is the total cutting time corresponding to the cutting path sequence.

[0066] For each contour geometric loop in the path cutting sequence, read the coordinates of its minimum bounding rectangle boundary. For two adjacent contour geometric loops, use the Euclidean distance formula to extract the Euclidean distance between all boundary point pairs (i.e., any two minimum bounding rectangle boundary points of the two contour geometric loops are considered as a boundary point pair), and take the minimum value as the minimum Euclidean distance of the contour boundary. ;

[0067] Furthermore, the effective cutting time and the idle travel time from the start to the corresponding contour geometric ring are read and added together to obtain the cutting time interval corresponding to that contour geometric ring. And combined with the preset material thermal diffusivity critical coefficient Extract path heat accumulation value ,Right now: ;

[0068] The total cutting time and path heat accumulation value of the workpiece to be cut are defined as the main optimization objective and the cooperative constraint objective of the workpiece to be cut, respectively. Specifically, an objective function is created with the cutting path sequence as the variable, denoted as... (i.e., the sum of cumulative idle travel time and effective cut-off time), the optimization direction is clearly minimized, that is... Any modification to the path (such as adjusting the contour cutting order) should be primarily evaluated based on whether it shortens the path. At the same time, a constraint function is created that takes the same sequence of cutting paths as variables, denoted as... This function quantifies the thermal impact risk of a path, and its specific calculation method is shown in the path thermal accumulation value. The constraints of the path thermal accumulation value are formally defined as inequalities. ,in, It is a dynamic constraint threshold;

[0069] The main optimization objective and the cooperative constraint objective of the workpiece to be cut are integrated to construct a path constraint optimization model for the workpiece to be cut, which is as follows: ,in, For decision variables, represents a sequence of cutting paths to be evaluated. The objective function, namely the total cutting time, is what needs to be minimized. The constraint is that the cumulative path heat value must not exceed the currently set dynamic threshold. This is the key mechanism for the model to achieve multi-objective collaboration. The feasible region represents the set of all possible cutting sequences that the path must belong to. This implies basic requirements such as the path needing to be a complete arrangement of all contours.

[0070] The specific steps to obtain the Pareto elite solution set of the workpiece to be cut are as follows: Mark the initial cutting path of the workpiece as the current cutting path, and set the constraint range and the Pareto empty set. Specifically: set the parameters for dynamic constraint iteration, set a dynamically changing threshold range for the path heat accumulation value in the cooperative constraint objective, define the starting value, ending value, and adjustment step size of the threshold. Simultaneously, create an empty set named the Pareto empty set to store all non-dominant Pareto optimal paths discovered during the optimization process. Set the initial cutting path as the starting path of the current search, denoted as... ;

[0071] Based on the set constraint range and the path constraint optimization model of the workpiece to be cut, an inner-layer local iterative search is performed on the current cutting path to update the current cutting path, and the Pareto empty set is updated simultaneously. Specifically:

[0072] Start a loop, starting the current constraint threshold from the initial value and gradually changing it according to the adjustment step size until the termination value is reached or exceeded. For each determined current constraint threshold, perform an inner optimization process once.

[0073] The current path obtained from inner layer optimization Starting with, execute a function designed to minimize the total cutting time. The local search process must satisfy constraints. For the current path Apply a pre-defined "path fragment reassembly operation" to generate a new path. In this embodiment, the preferred path segment recombination operation is the "2-opt" operation, which randomly selects two positions in the path sequence and reverses the order of all contours between these two positions, thereby generating a new path with a different structure but complete.

[0074] Calculate the new path The path heat accumulation value, if it is higher than If the new path violates the current constraints, it is discarded; otherwise, the new path is feasible, and its total cutting time is recalculated. ;

[0075] Compare the total cutting time of the feasible new path with that of the current path. If the total cutting time of the new path is shorter than that of the current path, then a better feasible path has been found, and it is accepted. As the new current path, that is to say Otherwise, the original current path will be retained. ;

[0076] Repeat the process of neighborhood generation, feasibility assessment and evaluation, and acceptance criteria until the inner convergence condition is met. The inner convergence condition can be set as follows: after multiple consecutive attempts (e.g., 50 times) to generate a new path, no path better than the current one can be found. Better (i.e.) If a smaller feasible path is found, or the preset maximum number of inner iterations is reached, then it is considered to be within the current constraint threshold. Below, a locally optimal feasible path has been found;

[0077] The current path obtained after inner layer optimization. (i.e., the corresponding threshold) The local optimal feasible path is compared with all existing paths in the Pareto empty set. The comparison is based on the Pareto dominance relation: if a path A is superior to path B in at least one objective and not inferior to path B in another objective, then A is said to dominate B.

[0078] If the current path If a path dominates a path in the solution set, then the dominated path is removed from the solution set. If it is not dominated by any path in the solution set, then it will Add to the solution set; if the current path If a path dominates a solution set, it will not be added to the solution set.

[0079] When the set constraint range reaches the preset termination condition, the updated Pareto empty set is marked as the Pareto elite solution set of the workpiece to be cut. Specifically, after completing one inner-layer optimization and solution set update, the dynamic constraint threshold is updated according to the adjustment step size, and the above steps are repeated until... When the value exceeds the preset range (i.e., the loop termination condition is met), the final Pareto elite solution set maintained when the outer loop ends is the output of this step. This set contains the most efficient (shortest total cutting time) path achievable under various heat accumulation constraint levels.

[0080] In this implementation plan, a constrained optimization model based on total cutting time and path heat accumulation value is constructed to accurately quantify efficiency and thermal impact risk. The model parameter extraction relies on the contour dataset mentioned above, with a closed-loop logic that fits the actual process. Dynamic constraint iteration combined with 2-opt local search is used to discover local optimal solutions under different heat accumulation thresholds. The Pareto elite solution set is updated simultaneously to include non-dominated solutions under multiple constraint levels, avoiding the limitations of a single optimal solution. Based on the screening rule of Euclidean distance of ideal points, the optimal path that balances efficiency and thermal stability can be quickly located, providing accurate and reliable optimized path support for subsequent processing.

[0081] Specifically, the steps for generating the optimized verification cutting path for the workpiece to be cut are as follows: perform multi-target verification processing on the optimized cutting path of the workpiece to be cut; and determine whether the optimized cutting path of the workpiece to be cut meets the preset verification qualification conditions, which specifically involves: traversing each closure accuracy value in the contour closure accuracy set, and judging one by one whether it meets the preset first qualification standard, that is, the closure accuracy value of the contour geometric loop must not exceed the preset contour closure accuracy threshold (such as 0.001); if the closure accuracy values ​​of all contours meet the above first qualification standard, then the geometric accuracy verification of the path is determined to be qualified.

[0082] Traverse each interference status identifier in the interference status identifier set and determine whether it meets the preset second qualification standard, that is, the interference status identifier of the empty travel path segment is 0; if the interference status identifier of all empty travel path segments is 0, then the empty travel safety check of the path is deemed qualified.

[0083] If the optimized cutting path of the workpiece to be cut meets the preset verification conditions (i.e., simultaneously meets the geometric accuracy verification and the no-travel safety verification), then the optimized cutting path of the workpiece to be cut is marked as an optimized verification cutting path; if the optimized cutting path of the workpiece to be cut does not meet the preset verification conditions, then the optimized cutting path of the workpiece to be cut is corrected to generate an optimized verification cutting path for the workpiece to be cut, specifically as follows:

[0084] If the geometric accuracy check fails, all contour geometric loops with excessive closure accuracy values ​​are traversed and marked as corrected contours. For each corrected contour, the sequence of cutting trajectory points in the current optimized cutting path is extracted. A third-order Bézier curve fitting algorithm is used to globally smooth and refit the original cutting trajectory point sequence of the contour, generating a new, smoother cutting trajectory with smaller beginning and end closure errors. The old trajectory of the corresponding contour in the original optimized cutting path is replaced with the newly generated cutting trajectory.

[0085] For empty travel safety verification failures, each empty travel path segment with status flag 1 is processed sequentially according to the path order. For the current empty travel segment to be corrected, its start and end points are fixed endpoints, and the rectangles in the "cut and occupied areas" set that trigger the interference judgment of this segment are used as obstacles. The A* path search algorithm is called to replan a collision-free obstacle avoidance path in the two-dimensional plane. The newly planned path must meet the additional constraint: its total length must not exceed 105% of the original empty travel segment length (i.e., path growth tolerance ≤ 5%). The corresponding empty travel segment in the original optimized cutting path is replaced by the newly planned safe path.

[0086] The revised path is used as the path to be verified, and the multi-target verification process is performed again until the generated path is determined to fully meet all the qualification conditions after a complete re-verification. The latest path is taken as the final result after the revision process is completed, and it is used as the optimized verification cutting path for the workpiece to be cut.

[0087] The specific steps for multi-target verification processing are as follows: Based on the contour geometry dataset of the workpiece to be cut, perform closure analysis on the optimized cutting path of the workpiece to be cut, and extract the contour closure accuracy set of the workpiece to be cut. Specifically, read the node coordinate sequence of each (arranged in the cutting order) contour geometry loop of the optimized cutting path of the workpiece to be cut, extract the trajectory start point and trajectory end point of each (arranged in the cutting order) contour geometry loop, and extract the distance value between the trajectory start point and trajectory end point of the contour geometry loop based on the Euclidean distance formula. Then, perform ratio processing with the total contour path length of the corresponding contour geometry loop in the contour geometry dataset to extract the closure accuracy value of each (arranged in the cutting order) contour geometry loop, i.e., the contour closure accuracy set.

[0088] Based on the contour geometry dataset of the workpiece to be cut, the optimized cutting path of the workpiece to be cut is processed by ray interferometry detection to extract the interference state identifier set of the workpiece to be cut. Specifically, an empty set is created to store the "cut and occupied area". The cutting process is simulated according to the contour cutting order of the optimized cutting path. After each contour cutting is simulated, the minimum bounding rectangle boundary coordinates corresponding to the contour are immediately obtained from the contour geometry dataset. This rectangular area is taken as a cut and occupied area and added to the created set.

[0089] After completing the cutting simulation of a contour, interference verification is performed on the subsequent empty travel path segments. For the current empty travel segment to be verified, all feature points on its path (including the start point, end point, and all turning points) are extracted. For each feature point, the ray method is performed: an infinitely long horizontal ray is emitted from the point in the positive X-axis direction; the total number of intersections between this ray and all rectangular areas in the currently cut and occupied area set is calculated. If the total number of intersections of the ray emitted by any feature point is odd, the empty travel segment is determined to have an interference risk, and the interference status is marked as 1; if the total number of intersections of the rays corresponding to all feature points is even, the empty travel segment is determined to be safe and interference-free, and the interference status is marked as 0. In this way, the interference status marks of several empty travel path segments of the workpiece to be cut are obtained.

[0090] In this implementation plan, contour closure accuracy and idle travel interference are the core verification dimensions. The path qualification is accurately determined by quantification thresholds and status indicators, covering key processing risk points. For unqualified paths, a third-order Bézier curve fitting is used to correct geometric deviations to ensure that the contour closure accuracy meets the standard. The obstacle avoidance idle travel is planned by the A* algorithm, and the path growth is strictly controlled to ≤5%, balancing safety and efficiency. The closed-loop verification and correction mechanism can be iterated repeatedly until the path is completely qualified. The final output optimized verification path has no accuracy deviation and no interference risk, which greatly reduces the probability of processing failure and improves cutting reliability.

[0091] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0092] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method of optimizing a laser cutting path of a laser cutting apparatus, characterized by, The method comprises the following steps: Obtain the geometric data of the workpiece to be cut, and perform contour analysis processing to extract the contour geometric data set of the workpiece to be cut; Based on the genetic algorithm, the contour geometric data set of the workpiece to be cut is subjected to cutting sequence planning processing to generate an initial cutting path of the workpiece to be cut; The initial cutting path of the workpiece to be cut is subjected to multi-objective collaborative optimization processing to obtain an optimized cutting path of the workpiece to be cut; The optimized cutting path of the workpiece to be cut is subjected to verification processing to generate an optimized verification cutting path of the workpiece to be cut.

2. The laser-cutting-path optimization method of a laser-cutting apparatus according to claim 1, characterized by, The specific steps of extracting the contour geometric data set of the workpiece to be cut are as follows: Read the geometric data of the workpiece to be cut, and perform identification processing to obtain a plurality of contour geometric rings of the workpiece to be cut; The space containing processing and the geometric feature extraction processing are performed on each contour geometric ring of the workpiece to be cut to obtain the ring ID, the nesting level and the geometric feature parameter set of each contour geometric ring of the workpiece to be cut.

3. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 2, wherein, The specific steps of obtaining the plurality of contour geometric rings of the workpiece to be cut are as follows: Read the geometric data of the workpiece to be cut, and perform preprocessing; Based on the preprocessed geometric data of the workpiece to be cut, a graph connection graph of the workpiece to be cut is constructed; Based on the depth-first search method, the graph connection graph of the workpiece to be cut is subjected to loop detection processing to generate a plurality of contour geometric rings of the workpiece to be cut.

4. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 2, wherein, The specific steps of generating the initial cutting path of the workpiece to be cut are as follows: Based on the ring ID of each contour geometric ring of the workpiece to be cut, an initial path population of the workpiece to be cut is generated, including a plurality of candidate initial paths, and each candidate initial path is formed by randomly associating and splicing each contour geometric ring; Based on the geometric feature parameter set of each contour geometric ring of each candidate initial path of the workpiece to be cut, an adaptability model of each candidate initial path of the workpiece to be cut is constructed, and the path adaptability score value of the corresponding candidate initial path is extracted; Based on the path adaptability score value of each candidate initial path of the workpiece to be cut, the initial path population of the workpiece to be cut is subjected to iteration processing until the preset convergence condition is met, and the initial cutting path of the workpiece to be cut is generated.

5. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 4, wherein, The specific steps of constructing the adaptability model of each candidate initial path of the workpiece to be cut are as follows: Based on the geometric feature parameter set of each contour geometric ring of each candidate initial path of the workpiece to be cut, the path evaluation data set of the corresponding candidate initial path is extracted, including the idle stroke value and the nesting conflict frequency value; Based on the path evaluation data set of each candidate initial path of the workpiece to be cut, the adaptability model of the corresponding candidate initial path is constructed.

6. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 1, wherein, The specific steps of obtaining the optimized cutting path of the workpiece to be cut are as follows: Based on the initial cutting path and the contour geometric data set of the workpiece to be cut, a path constraint optimization model of the workpiece to be cut is constructed; The path constraint optimization model of the workpiece to be cut is subjected to dynamic constraint iteration optimization processing to obtain a Pareto elite solution set of the workpiece to be cut; Based on the preset decision rule, the Pareto elite solution set of the workpiece to be cut is subjected to screening processing to generate an optimized cutting path of the workpiece to be cut.

7. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 6, wherein, The specific steps of constructing the path constraint optimization model of the workpiece to be cut are as follows: Based on the initial cutting path of the workpiece to be cut and the profile geometry data set, the total cutting time and the path heat accumulation value of the workpiece to be cut are extracted respectively; The total cutting time and the path heat accumulation value of the workpiece to be cut are defined as the main optimization target and the collaborative constraint target of the workpiece to be cut respectively; The main optimization target and the collaborative constraint target of the workpiece to be cut are integrated to construct the path constraint optimization model of the workpiece to be cut.

8. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 6, wherein, The specific steps of obtaining the Pareto elite solution set of the workpiece to be cut are as follows: The initial cutting path of the workpiece to be cut is marked as the current cutting path, and the constraint range and the Pareto empty set are set; Based on the set constraint range and the path constraint optimization model of the workpiece to be cut, the current cutting path is processed by inner local iteration search to update the current cutting path, and the Pareto empty set is updated synchronously; When the set constraint range reaches the preset termination condition, the updated Pareto empty set is marked as the Pareto elite solution set of the workpiece to be cut.

9. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 1, wherein, The specific steps of generating the optimized verification cutting path of the workpiece to be cut are as follows: The optimized cutting path of the workpiece to be cut is processed by multi-objective verification; And judge whether the optimized cutting path of the workpiece to be cut meets the preset verification qualified condition; If the optimized cutting path of the workpiece to be cut meets the preset verification qualified condition, the optimized cutting path of the workpiece to be cut is marked as the optimized verification cutting path; If the optimized cutting path of the workpiece to be cut does not meet the preset verification qualified condition, the optimized cutting path of the workpiece to be cut is processed by correction to generate the optimized verification cutting path of the workpiece to be cut.

10. The method of optimizing a laser cutting path of a laser cutting apparatus according to claim 9, wherein, The specific steps of multi-objective verification processing are as follows: Based on the profile geometry data set of the workpiece to be cut, the optimized cutting path of the workpiece to be cut is analyzed by closure to extract the profile closure precision set of the workpiece to be cut; Based on the profile geometry data set of the workpiece to be cut, the optimized cutting path of the workpiece to be cut is processed by ray method interference detection to extract the interference state identification set of the workpiece to be cut.

Images