A method and system for optimizing a laser cutting path for a laser

Abstract

The present application relates to the technical field of data processing, and more particularly to a laser cutting path optimization method and system. The method comprises the following steps: collecting the cutting points of the to-be-cut plate, obtaining the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in each iteration, and obtaining the objective function value of the new solution in each iteration; according to the objective function value of the new solution in each iteration and the primary destruction degree, obtaining the destruction degree of the random removal operator in the ALNS algorithm when generating the new solution in each iteration; according to the destruction degree of the random removal operator in the ALNS algorithm when generating the new solution in each iteration, obtaining the destruction proportion of the random removal operator in the ALNS algorithm when generating the new solution in each iteration; and according to the destruction proportion of the random removal operator in the ALNS algorithm when generating the new solution in each iteration, combining the SA algorithm and the ALNS algorithm to obtain the optimal path of the to-be-cut plate. The present application improves the accuracy of the optimal path of the to-be-cut plate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a laser cutting path optimization method and system. BACKGROUND

[0002] Under the background of the current intelligent upgrading of manufacturing industry, laser cutting technology, as a core means of high-precision processing, is applied in the fields of automobiles, aerospace, consumer electronics, etc. However, the traditional path planning method often lacks flexibility, which may cause unnecessary backtracking or redundant path, thereby wasting time and energy and reducing cutting efficiency. The present scheme introduces the SA algorithm to complete the optimization of the laser cutting path of the laser, which can jump out of the local optimum and efficiently search for the global optimal path in the solution space through the mechanism of probabilistically accepting inferior solutions combined with the temperature decay iterative search strategy. The application of the algorithm meets the needs of flexible manufacturing and energy saving in the era of Industry 4.0, and can directly shorten the processing time and reduce the production cost of enterprises.

[0003] However, the SA algorithm only generates a new solution at each iteration within the neighborhood range of the current solution, which cannot effectively explore the entire search space, especially for complex cutting paths, which may cause the SA algorithm to fall into a local optimal solution, resulting in a slow convergence speed in the later iteration period and an inability to accurately find the optimal solution. Therefore, the ALNS algorithm (adaptive large neighborhood search algorithm) is used to generate a new solution at each iteration to accelerate the generation of high-quality solutions by dynamically selecting the destruction operator and the repair operator.

[0004] Since the destruction operator removes the cutting points in the current solution at a fixed destruction ratio, which breaks the local structure of the current solution, if the destruction ratio is set too large, it may cause excessive disturbance in the later iteration period, making the algorithm unable to effectively converge to the optimal solution. If the destruction ratio is too small, the search in the early iteration period will be too limited, which cannot effectively explore the solution space, causing the algorithm to easily stay in the local optimal region and miss the global optimal solution. SUMMARY

[0005] In order to solve the technical problem that the destruction ratio of the destruction operator is too small or too large and the optimal solution cannot be accurately obtained, the present application provides a laser cutting path optimization method and system.

[0006] In the first aspect, the present application provides a laser cutting path optimization method, which adopts the following technical scheme:

[0007] A laser cutting path optimization method, comprising the steps of:

[0008] Collecting the cutting points of the plate to be cut;

[0009] In the process of the SA algorithm iteratively obtaining the optimal path of the to-be-cut plate, according to the number of cutting points of the to-be-cut plate and the temperature value of the SA algorithm when a new solution is generated in each iteration, the primary damage degree of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration is obtained; the objective function value of the new solution generated in each iteration is obtained;

[0010] According to the objective function value of the new solution generated in each iteration and the primary damage degree, the damage degree of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration is obtained;

[0011] According to the damage degree of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration, the damage proportion of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration is obtained; and the optimal path of the to-be-cut plate is obtained by combining the SA algorithm and the ALNS algorithm according to the damage proportion of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration.

[0012] The innovation of the present application lies in that the random removal operator in the adaptive large neighborhood operation (ALNS) is introduced into the SA algorithm, the primary damage degree of the random removal operator in the ALNS algorithm when a new solution is generated in each iteration is obtained according to the number of cutting points of the to-be-cut plate and the temperature value of the SA algorithm when the new solution is generated in each iteration, then the initial damage degree is adjusted according to the quality of the new solution generated in each iteration to obtain the damage degree of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration, and further obtain the damage proportion of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration. By dynamically adjusting the damage proportion of the random removal operator, the disturbance intensity can be flexibly controlled according to the quality and damage degree of the current solution in different iteration stages, so as to search in a larger solution space, so that the optimal path can be found more quickly and accurately in the complex laser cutting path problem.

[0013] Preferably, the primary damage degree of the random removal operator in the ALNS algorithm when a new solution is generated in each iteration comprises:

[0014] ;

[0015] In the formula, represents the primary damage degree of the random removal operator in the ALNS algorithm when a new solution is generated in the ith iteration; represents the number of cutting points of the to-be-cut plate; represents the temperature value of the SA algorithm when a new solution is generated in the ith iteration; represents the initial temperature value of the preset SA algorithm; and norm() represents a normalization function.

[0016] The number of cutting points of the plate to be cut and the temperature value of the SA algorithm are considered to obtain the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in each iteration.

[0017] Preferably, the target function value when generating a new solution in each iteration comprises:

[0018] ;

[0019] In the formula, represents the target function value when generating a new solution in the i th iteration; represents the path length in the entire cutting process when the laser cuts the same plate as the plate to be cut according to the new solution generated in the i th iteration; represents the temperature increment in the entire cutting process when the laser cuts the same plate as the plate to be cut according to the new solution generated in the i th iteration; exp() represents an exponential function with a natural constant as the base; and norm() represents a normalization function.

[0020] reflects the quality of generating a new solution in each iteration.

[0021] Preferably, the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in each iteration comprises:

[0022] ;

[0023] In the formula, represents the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in the i th iteration; represents the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in the i th iteration; represents the number of historical iterations before the i th iteration; represents the target function value of the new solution generated in the m th historical iteration before the i th iteration; represents the target function value of the current solution of the m th historical iteration before the i th iteration; represents the range of the target function values of all historical iterations before the i th iteration; || represents an absolute value symbol; norm() represents a normalization function; and exp() represents an exponential function with a natural constant as the base; represents a preset adjustment coefficient.

[0024] Preferably, the number of historical iterations before the i th iteration comprises:

[0025] The number of historical iterations before each iteration is set to M, and the M iterations before each iteration are regarded as the historical iterations before the i th iteration.

[0026] The damage degree of the random removal operator in the ALNS algorithm for generating a new solution in each iteration is acquired.

[0027] Preferably, the damage ratio of the random removal operator in the ALNS algorithm for generating a new solution in each iteration is acquired, and the damage ratio of the random removal operator in the ALNS algorithm for generating a new solution in each iteration is acquired.

[0028]

[0029] In the formula, i represents the i-th iteration for generating a new solution; and represents the damage ratio of the random removal operator in the ALNS algorithm for generating a new solution in the i-th iteration. In the formula, i represents the i-th iteration for generating a new solution; and represents the damage ratio of the random removal operator in the ALNS algorithm for generating a new solution in the i-th iteration. In the formula, i represents the i-th iteration for generating a new solution; and represents the damage degree of the random removal operator in the ALNS algorithm for generating a new solution in the i-th iteration. In the formula, represents the damage ratio of the random removal operator in the preset ALNS algorithm.

[0030] The disturbance intensity can be flexibly controlled according to the quality and damage degree of the current solution in different iteration stages.

[0031] Preferably, the acquisition of the temperature value of the SA algorithm for generating a new solution in the i-th iteration comprises the following steps.

[0032] A preset temperature cooling coefficient is provided, the temperature value of the SA algorithm for generating a new solution in the first iteration is an initial temperature value, the temperature value of the SA algorithm for generating a new solution in the second iteration is a product of the temperature value of the SA algorithm for generating a new solution in the first iteration and the temperature cooling coefficient, and the temperature value of the SA algorithm for generating a new solution in the i-th iteration is a product of the temperature value of the SA algorithm for generating a new solution in the i-1th iteration and the temperature cooling coefficient.

[0033] Preferably, the acquisition of the cutting point of the to-be-cut plate comprises the following steps.

[0034] The CAD drawing file of the to-be-cut plate is input into an API interface of CAD software for loading, and the position coordinates of all cutting points in the to-be-cut plate are extracted.

[0035] Preferably, the acquisition of the range of the objective function values of all historical iterations for generating a new solution before the i-th iteration comprises the following steps.

[0036] The difference between the maximum value and the minimum value of the objective function values of all historical iterations for generating a new solution before the i-th iteration is taken as the range of the objective function values of all historical iterations for generating a new solution before the i-th iteration.

[0037] In the second aspect, the application provides a laser cutting path optimization system for a laser, and adopts the following technical scheme.

[0038] ​The laser laser cutting path optimization system comprises a processor and a memory, and the memory stores computer program instructions which, when executed by the processor, implement the above-mentioned laser laser cutting path optimization method.

[0039] By adopting the above technical solution, the above-mentioned laser laser cutting path optimization method is generated into a computer program and stored in the memory to be loaded and executed by the processor, so that a terminal device is manufactured according to the memory and the processor, and use is facilitated.

[0040] The present application has the following technical effects: by introducing the random removal operator in the adaptive large neighborhood operation (ALNS) into the SA algorithm, according to the number of cutting points of the cutting plate, the temperature value of the SA algorithm when generating a new solution each time, and the quality of the new solution generated each time, the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution each time is obtained, and based on the destruction degree, the destruction proportion of the random removal operator in the ALNS algorithm when generating a new solution each time is adaptively obtained, so that the disturbance intensity can be flexibly controlled according to the quality and destruction degree of the current solution at different iteration stages, and the optimal path can be found more quickly and accurately in a complex laser cutting path problem. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 is a method flowchart in a laser laser cutting path optimization method of an embodiment of the present application. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments of the present application.

[0043] An embodiment of the present application discloses a laser laser cutting path optimization method, referring to Figure 1 , comprising steps S1-S4:

[0044] S1: Collecting cutting points of a cutting plate.

[0045] In the embodiment of the present application, the CAD drawing file of the cutting plate is input into the API interface of the CAD software for loading, and the position coordinates of all cutting points in the cutting plate are extracted; wherein the cutting point usually refers to a point at a division or boundary in the plate, and the position coordinates of the endpoints or intersection points of all lines or multi-segment lines are obtained through the API.

[0046] S2: Obtaining the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution each time.

[0047] It should be noted that the application outputs the optimal path when cutting the plate by the laser according to the SA algorithm, the SA algorithm (simulated annealing algorithm) usually obtains the optimal solution after multiple iterations, has a current solution at each iteration, and generates a new solution according to the current solution, then selects one of the current solution and the new solution at each iteration, and selects one of the current solution and the new solution at each iteration. Good solution as the current solution of the next iteration, and so on until the iteration is stopped after reaching the iteration times, and the current solution of the last iteration is the optimal solution; however, when the SA algorithm obtains the optimal cutting path, each iteration only makes small changes (such as exchanging two cutting points in the cutting path) in the neighborhood of the current solution to generate a new solution at each iteration. This operation may not effectively explore the entire search space, especially for complex cutting paths. This limitation may cause the SA algorithm to fall into a local optimal solution, resulting in slow convergence in the later iteration period and inability to accurately find the optimal solution.

[0048] Therefore, the ALNS algorithm (adaptive large neighborhood search algorithm) is used at each iteration to generate a new solution at each iteration, and the dynamic selection of the destruction operator (destroying the structure of the current solution) and the repair operator (reconstructing the solution according to the destroyed structure) is used to accelerate the generation of high-quality solutions. However, one of the destruction operators, a random removal operator, removes the cutting points in the current solution at a fixed destruction ratio, thereby breaking the local structure of the current solution. Then, the repair operator reconstructs the complete solution based on the broken structure, i.e. obtains a new solution at each iteration.

[0049] Although this operation introduces more changes by breaking the local structure to avoid the algorithm falling into a local optimum, the selection of the same destruction ratio will affect the convergence of the algorithm. If the destruction ratio is set to be large, it may cause excessive disturbance (i.e. large changes to the current solution) in the later iteration period. Such disturbance may destroy the solution that is close to the optimal solution, making the algorithm unable to effectively converge to the optimal solution. If the destruction ratio is too small, the search in the early iteration period will be too limited, and the solution space cannot be effectively explored, resulting in the algorithm easily staying in the local optimal region and missing the global optimal solution. Therefore, the application analyzes the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution at each iteration, and then adaptively sets the destruction ratio of the random removal operator in the ALNS algorithm when generating a new solution at each iteration based on the destruction degree, so as to ensure that the optimal solution quality is guaranteed while the algorithm calculation efficiency is improved as much as possible.

[0050] It needs to be further explained that the more the number of cutting points of the to-be-cut plate material, the larger the problem scale faced when obtaining the cutting path of the to-be-cut plate material, and then the larger the destruction ratio needs to be set for the random removal operator in the ALNS algorithm to randomly remove part of the cutting points when generating a new solution each time, and at this time, the primary destruction degree of the random removal operator in the ALNS algorithm is larger each time when generating a new solution; and in the SA algorithm, when the temperature value of the SA algorithm approaches the initial temperature, it indicates that the algorithm is in the early stage of search, and at this time, a larger disturbance is needed to explore the solution space in a large range, and therefore, according to the number of cutting points of the to-be-cut plate material and the temperature value of the SA algorithm when generating a new solution each time, the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution each time is obtained.

[0051] In the embodiment of the application, the initial temperature value of the preset SA algorithm is 100℃, the temperature threshold value is 1℃, and the temperature cooling coefficient is 0.98, and the SA algorithm is used to iteratively obtain the optimal solution (optimal path) of the to-be-cut plate material; wherein the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution each time is:

[0052] ;

[0053] In the formula, represents the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution for the i th time; represents the number of cutting points of the to-be-cut plate material; represents the temperature value of the SA algorithm when generating a new solution for the i th time; represents the initial temperature value of the preset SA algorithm, and the preset is in Celsius; norm() represents a normalization function;

[0054] It needs to be noted that the temperature value of the SA algorithm when generating a new solution for the first time is the initial temperature value, the temperature value of the SA algorithm when generating a new solution for the second time is the product of the temperature value of the SA algorithm when generating a new solution for the first time and the temperature cooling coefficient, and the temperature value of the SA algorithm when generating a new solution for the third time is the product of the temperature value of the SA algorithm when generating a new solution for the second time and the temperature cooling coefficient;

[0055] The larger the value of is, the more the number of cutting points of the to-be-cut plate material, and the larger the problem scale faced when obtaining the optimal cutting path of the to-be-cut plate material, and then the larger the destruction ratio needs to be set for the random removal operator in the ALNS algorithm to randomly remove part of the cutting points each time when generating a new solution, and at this time, the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution for the i th time is larger;

[0056] Since the temperature value of the SA algorithm will be continuously reduced with the increase of the iteration number, the iteration stops when the temperature threshold is reached, therefore The greater the value is, the closer the temperature value of the SA algorithm in the generation of the new solution of the i-th iteration is to the initial temperature value, the greater the destruction ratio required by the i-th iteration is, and the greater the primary destruction degree of the random removal operator in the ALNS algorithm in the generation of the new solution of the i-th iteration is.

[0057] S3: obtaining the objective function value of the new solution generated by each iteration; obtaining the destruction degree of the random removal operator in the ALNS algorithm in the generation of the new solution of each iteration according to the objective function value of the new solution generated by each iteration and the primary destruction degree; and obtaining the destruction ratio of the random removal operator in the ALNS algorithm in the generation of the new solution of each iteration according to the destruction degree of the random removal operator in the ALNS algorithm in the generation of the new solution of each iteration.

[0058] It should be noted that, in the whole process of cutting the plate by the laser according to the arbitrary cutting path, the greater the temperature increment of cutting the plate is, the greater the possibility of causing thermal stress distortion of the cutting path is, that is, the greater the possibility of causing the deformation of the plate by the cutting path is, so the cutting path is less likely to be the optimal path, the greater the objective function value of the cutting path is, and the smaller the path length in the whole process of cutting the plate by the laser according to the cutting path is, the higher the working efficiency corresponding to the cutting path is, the more likely the cutting path is the optimal path, and the smaller the objective function value of the cutting path is, so the objective function value of the cutting path is obtained according to the temperature increment and the path length in the whole process of cutting the plate by the laser according to the arbitrary cutting path, the greater the objective function value is, the lower the quality of the cutting path is, and the smaller the objective function value is, the higher the quality of the cutting path is.

[0059] In the embodiment of the present application, the objective function value of the new solution generated by each iteration is obtained as follows:

[0060] ;

[0061] In the formula, represents the objective function value of the new solution generated by the i-th iteration; represents the path length in the whole cutting process when the laser cuts the same plate as the plate to be cut according to the new solution generated by the i-th iteration; represents the temperature increment in the whole cutting process when the laser cuts the same plate as the plate to be cut according to the new solution generated by the i-th iteration; exp() represents the exponential function with the natural constant as the base; and norm() represents the normalization function.

[0062] It should be noted that the temperature of the SA algorithm and the number of cutting points (problem size) are used to preliminarily set the destruction ratio in the ALNS algorithm, which considers the general phased demand, such as a higher temperature in the early stage requiring a larger destruction ratio, and a lower temperature in the later stage reducing the destruction ratio, but the disadvantage of this setting method is that it only depends on the temperature change of the algorithm and the size of the problem, and ignores the quality of the solution in each iteration, assuming that the temperature of the SA algorithm has been reduced, the search range of the algorithm may also be too limited to find a better solution, resulting in falling into a local optimal solution, so the quality of the solution generated in each iteration in the SA algorithm is used to obtain the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in each iteration;

[0063] It should be further pointed out that if the quality of the new solution generated in the previous several iterations in the iteration of the SA algorithm is poor or changes little, the destruction ratio of the ALNS algorithm should be increased in the current iteration to guide the search to a wider solution space, and the greater the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in the current iteration.

[0064] In the embodiment of the application, the number of historical iterations before each iteration is preset as M=5, and the M iterations before each iteration are regarded as the historical iterations before the ith iteration;

[0065] The destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in each iteration is obtained:

[0066] ;

[0067] In the formula, represents the destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in the ith iteration; represents the primary destruction degree of the random removal operator in the ALNS algorithm when generating a new solution in the ith iteration; represents the number of historical iterations before the ith iteration; represents the objective function value of the new solution generated in the mth historical iteration before the ith iteration; represents the objective function value of the current solution of the mth historical iteration before the ith iteration; represents the range of the objective function values of all the historical iterations before the ith iteration; || represents the absolute value symbol; norm() represents the normalization function; exp() represents the exponential function with the natural constant as the base number; represents a preset adjustment coefficient, which is used to adjust the value range of to [0.5-1.5];

[0068] The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration.

[0069] The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. The greater the value of the i-th iteration, the greater the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, and at this time, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration.

[0070] It should be noted that if the number of historical iterations before the i-th iteration does not satisfy five times, the primary destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration is directly taken as the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration.

[0071] It should be noted that the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration, the greater the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration.

[0072] In the embodiment of the present application, the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in each iteration is obtained as follows:

[0073]

[0074] In the formula, i represents the i-th iteration, and represents the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. In the formula, i represents the i-th iteration, and represents the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. In the formula, i represents the i-th iteration, and represents the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. In the formula, i represents the i-th iteration, and represents the destruction degree of the random removal operator in the ALNS algorithm when the new solution is generated in the i-th iteration. In the embodiment of the present application, the preset destruction degree of the random removal operator in the ALNS algorithm is 25%, and in other embodiments, the implementation personnel can preset the destruction degree of the random removal operator in the ALNS algorithm according to the specific implementation. ​The greater the destruction degree of the random removal operator in the ALNS algorithm when a new solution is generated in the i-th iteration, the greater the destruction ratio of the random removal operator in the ALNS algorithm when a new solution is generated in the i-th iteration, and the local optimal solution is jumped out by exploring a larger range of solution space.

[0075] S4: The destruction ratio of the random removal operator in the ALNS algorithm when a new solution is generated in each iteration is combined with the SA algorithm and the ALNS algorithm to obtain the optimal path of the to-be-cut plate.

[0076] It should be noted that the destruction ratio of the random removal operator in the ALNS algorithm when a new solution is generated in each iteration is obtained, and the optimal solution (optimal path) of the to-be-cut plate is iteratively obtained by combining the SA algorithm and the ALNS algorithm.

[0077] In the embodiment of the application, the specific operation of iteratively obtaining the optimal solution (optimal path) of the to-be-cut plate by using the SA algorithm and the ALNS algorithm is as follows:

[0078] The initial temperature value of the preset SA algorithm is 100 DEG C, the temperature threshold is 1 DEG C, the temperature cooling coefficient is 0.98, the new solution acceptance probability is 10%, a path is randomly generated according to the cutting points of the to-be-cut plate as an initial solution, and the initial solution is taken as the current solution of the first iteration;

[0079] The steps of obtaining the ALNS algorithm after the first iteration optimization are as follows:

[0080] The destruction operators include a random removal operator (the cutting points in the current solution of the first iteration are randomly removed according to the destruction ratio of the random removal operator in the ALNS algorithm when a new solution is generated in the first iteration) and a distance-sensitive removal operator (in the current solution of the first iteration, the first 10% of the cutting points that contribute most to the path distance are removed); the repair operators include a random insertion operator (the cutting points removed by the destruction operator are randomly inserted into any position in the path) and a greedy insertion operator (the cutting points removed by the destruction operator are inserted into a position that makes the path distance increment small).

[0081] The first iteration score of each operator is set to 1, the candidate probability of each operator in the first iteration is obtained according to the first iteration score of each operator (the prior art step in the ALNS algorithm), and any one destruction operator and repair operator are selected by using a roulette wheel according to the candidate probability of each operator in the first iteration, which is recorded as the destruction operator and repair operator of the first iteration.

[0082] The current solution of the first iteration is destroyed and repaired by using the ALNS algorithm after the first iteration optimization to obtain a new solution generated in the first iteration; and the temperature of the SA algorithm in the first iteration is the initial temperature.

[0083] obtaining the objective function value of the new solution generated by the first iteration and the objective function value of the current solution of the first iteration, if the objective function value of the new solution generated by the first iteration is greater than the objective function value of the current solution of the first iteration, then determining whether to accept the new solution generated by the first iteration according to the acceptance probability of the new solution, if accepted, taking the new solution generated by the first iteration as the current solution of the second iteration, otherwise, taking the current solution of the first iteration as the current solution of the second iteration; if the objective function value of the new solution generated by the first iteration is less than the objective function value of the current solution of the first iteration, taking the new solution generated by the first iteration as the current solution of the second iteration;

[0084] The steps of the ALNS algorithm after the second round of iteration optimization are as follows:

[0085] The destruction operators include: a random removal operator (removing the cutting points in the current solution of the first iteration according to the destruction proportion of the random removal operator in the ALNS algorithm when the new solution of the second iteration is generated), and a distance-sensitive removal operator (removing the top 10% of the cutting points that contribute most to the path distance in the current solution of the second iteration); the repair operators include: a random insertion operator (randomly inserting the cutting points removed by the destruction operator into any position in the path), and a greedy insertion operator (inserting the cutting points removed by the destruction operator into the position with the smallest path distance increment);

[0086] If the objective function value of the new solution generated by the first iteration is less than the current solution of the first iteration, then the scores of the destruction operator and the repair operator of the first iteration are increased by one, if the objective function value of the new solution generated by the first iteration is greater than the objective function value of the current solution of the first iteration, and the new solution generated by the first iteration is taken as the current solution of the second iteration, then the scores of the destruction operator and the repair operator of the first iteration are increased by 0.5, to obtain the second iteration score of each operator;

[0087] According to the second iteration score of each operator, the candidate probability of each operator in the second round of iteration is obtained, and according to the candidate probability of each operator in the second round of iteration, a roulette is used to select any destruction operator and repair operator as the destruction operator and repair operator of the second iteration;

[0088] The current solution of the second iteration is destroyed and repaired according to the ALNS algorithm after the second round of iteration optimization, to obtain the new solution generated by the third iteration; the temperature of the second iteration SA algorithm is the product of the temperature of the second iteration SA algorithm and the temperature cooling coefficient;

[0089] Similarly, until the temperature of the SA algorithm in the Jth iteration is less than or equal to the temperature threshold, the current solution of the Jth iteration is taken as the optimal path of the cutting plate to be cut.

[0090] The above are preferred embodiments of the present application, and do not limit the protection scope of the present application, so: any equivalent changes made according to the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method for optimizing the laser cutting path, characterized in that, include: Collect the cutting points of the material to be cut; During the iterative acquisition of the optimal path for the material to be cut by the SA algorithm, the initial degree of destruction of the random removal operator in the ALNS algorithm is obtained based on the number of cutting points on the material and the temperature value of the SA algorithm when a new solution is generated in each iteration. This includes: ； In the formula, This represents the initial degree of destruction in the ALNS algorithm when a new solution is generated in the i-th iteration; This represents the number of cutting points on the material to be cut. This represents the temperature value of the SA algorithm when a new solution is generated in the i-th iteration; This represents the initial temperature value for the preset SA algorithm; norm() represents the normalization function. Obtain the objective function value for each iteration that generates a new solution, including: ； In the formula, This represents the objective function value used to generate a new solution in the i-th iteration. The path length during the entire cutting process represents the length of the laser when it generates a new solution based on the i-th iteration to cut a material identical to the material to be cut. exp() represents the temperature increment during the entire cutting process when the laser generates a new solution based on the i-th iteration and cuts a material identical to the material to be cut; exp() represents an exponential function with the natural constant as the base; norm() represents a normalization function. Based on the objective function value generated in each iteration and the initial degree of corruption, the degree of corruption of the random removal operator in the ALNS algorithm during each iteration is obtained, including: ； In the formula, This represents the degree of disruption caused by the random removal of operators in the ALNS algorithm when generating a new solution in the i-th iteration. This represents the initial degree of destruction in the ALNS algorithm when a new solution is generated in the i-th iteration; This represents the number of previous iterations before the i-th iteration; This represents the objective function value used to generate a new solution in the m-th historical iteration prior to the i-th iteration. The objective function value represents the current solution of the historical iteration before the i-th iteration; represents the range of the objective function values ​​generated by all previous iterations before the i-th iteration; || represents the absolute value sign; norm() represents the normalization function; exp() represents the exponential function with the natural constant as the base; This represents the preset adjustment coefficient; Based on the degree of destruction caused by the random removal of operators in the ALNS algorithm when generating a new solution in each iteration, and taking into account the requirements of different stages in the SA algorithm and the quality of the solution in each iteration, the destruction ratio of the random removal of operators in the ALNS algorithm when generating a new solution in each iteration is obtained. Based on the destruction ratio of the random removal of operators in the ALNS algorithm when generating a new solution in each iteration, the optimal path of the plate to be cut is obtained by combining the SA algorithm and the ALNS algorithm.

2. The laser cutting path optimization method according to claim 1, characterized in that, Obtaining the number of historical iterations prior to the i-th iteration includes: The number of previous iterations before each iteration is set to M. The M iterations before each iteration are taken as the previous iterations before the i-th iteration.

3. The laser cutting path optimization method according to claim 1, characterized in that, The process of obtaining the destruction ratio of randomly removed operators in the ALNS algorithm during each iteration to generate a new solution includes: ； In the formula, This represents the proportion of disruption caused by the random removal of operators in the ALNS algorithm when a new solution is generated in the i-th iteration; This represents the degree of disruption caused by the random removal of operators in the ALNS algorithm when generating a new solution in the i-th iteration. This represents the proportion of disruption caused by randomly removing operators in the default ALNS algorithm.

4. The laser cutting path optimization method according to claim 1, characterized in that, The acquisition of the temperature value of the SA algorithm when generating a new solution in the i-th iteration includes: A preset temperature cooling coefficient is used. When a new solution is generated in the first iteration, the temperature value of the SA algorithm is the initial temperature value. When a new solution is generated in the second iteration, the temperature value of the SA algorithm is the product of the temperature value of the SA algorithm when a new solution is generated in the first iteration and the temperature cooling coefficient. And so on. When a new solution is generated in the i-th iteration, the temperature value of the SA algorithm is the product of the temperature value of the SA algorithm when a new solution is generated in the (i-1)-th iteration and the temperature cooling coefficient.

5. The laser cutting path optimization method according to claim 1, characterized in that, The process of collecting the cutting points of the material to be cut includes: The CAD drawing file of the board to be cut is imported into the API interface of the CAD software for loading, and the position coordinates of all cutting points in the board to be cut are extracted.

6. The laser cutting path optimization method according to claim 1, characterized in that, The acquisition of the range of the objective function values ​​generated from all previous iterations before the i-th iteration includes: The difference between the maximum and minimum values ​​of the objective function values ​​generated by all previous iterations before the i-th iteration is taken as the range of the objective function values ​​generated by all previous iterations before the i-th iteration.

7. A laser cutting path optimization system, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement a laser cutting path optimization method according to any one of claims 1-6.

Images