Steel plate laser cutting path planning method and system based on step-by-step optimization

Abstract

The application provides a steel plate piece laser cutting path planning method and system based on step-by-step optimization, and relates to the technical field of production cutting. First, the cutting sequence of the steel plate piece is optimized by using a simulated annealing algorithm of adaptive large-range neighborhood search. Then, the cutting starting point of each steel plate piece is optimized by using a genetic algorithm. Finally, the cutting path is generated based on the cutting sequence of the steel plate piece and the cutting starting point of each steel plate piece. The application can reduce the cutting path, improve the operation efficiency, shorten the production cycle and improve the automation level of the production line.

Description

Technical Field

[0001] This invention relates to the field of production cutting technology, and in particular to a method and system for laser cutting path planning of steel plates based on step-by-step optimization. Background Technology

[0002] In industrial production, an order often requires a large number of steel plates of varying sizes. To reduce material waste, these steel plates need to be rationally arranged on a raw material base plate before the plates are cut from the base plate. In modern manufacturing with large-scale mass production, this method can reduce material waste and improve material utilization.

[0003] Cutting path planning refers to the use of algorithms and software to plan the optimal cutting path during cutting processes, in order to improve production efficiency and reduce waste. In industrial production, cutting path planning is a crucial step. Given a fixed cutting speed, the choice of path directly affects laser processing time; a well-designed cutting path can make the cutting process more efficient and precise.

[0004] Since path planning problems can be reduced to the Traveling Salesman Problem (TSP), which is NP-hard, it is difficult to find an exact solution. Intelligent algorithms are often used to solve path planning problems by numbering the steel plates and their vertices, performing a combined encoding, and then iteratively searching for possible paths according to a certain strategy, ultimately selecting the shortest path as the cutting path. However, in large-scale mass production, the number of steel plates to be cut at one time is large, and the number of vertices in the steel plates is uncertain, resulting in a very large solution space. Using a combined encoding approach leads to many invalid searches, resulting in long solution times, low solution quality, and ultimately, unreasonable path planning. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for laser cutting path planning of steel plates based on step-by-step optimization, which can reduce the cutting path, improve operating efficiency, shorten the production cycle and improve the level of production line automation.

[0006] A method for laser cutting path planning of steel plates based on step-by-step optimization, comprising:

[0007] S1, Determine the optimal cutting sequence for the steel plate:

[0008] S11, number each steel plate in the steel plate layout result diagram, abstract each steel plate as a point, and then calculate the distance matrix between all steel plates;

[0009] S12, based on the distance matrix, adopts an adaptive large-range neighborhood search simulated annealing algorithm, randomly generates the initial cutting order, encodes the numbers of the steel plates according to the initial cutting order, initializes the weights wDestroy and wRepair of the search operator, and initializes the scores destroyScore and repairScore of the search operator.

[0010] S13 uses a roulette wheel method, based on wDestroy and wRepair, to select a search operator, generate a new cutting order, update the order code using the Metropolice criterion, and score the new order code to update destroyScore and repairScore.

[0011] S14, update the weights wDe stroy and wRepair of the search operator based on the updated destroyScore and repairScore, and update the annealing temperature and the current optimal cutting order;

[0012] S15, repeat S13-S14 until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate;

[0013] S2, Determine the optimal cutting starting point for the steel plate:

[0014] S21, take all the vertices of each steel plate as a set, and number each vertex in the set;

[0015] S22. Using a genetic algorithm, the number of one vertex in each steel plate is selected according to the optimal cutting order of the steel plate and combined and encoded, which is recorded as the initial cutting starting point set. Multiple initial cutting starting point sets are randomly generated.

[0016] S23, For each initial cutting starting point set, use the exchange operator to generate a new cutting starting point set, and use the Metropolice criterion to update all initial cutting starting point sets;

[0017] S24, repeat S23 until the set number of iterations is reached to obtain the optimal cutting starting point set, that is, the optimal cutting starting point for each steel plate.

[0018] S3, the cutting path is obtained based on the optimal cutting sequence of the steel plates and the optimal cutting starting point of each steel plate.

[0019] Optionally, the neighborhood search of the simulated annealing algorithm for adaptive large-range neighborhood search is achieved by combining the Destroy operator and the Repair operator. The Destroy operator includes randomDestroy and maxDestroy; the Repair operator includes greedyinsert and randominsert; the initial weights of the Destroy operator and the Repair operator are set to be the same, and the scores of the Destroy operator (destroyScore) and the Repair operator (repairScore) are initialized to 0.

[0020] Optionally, S13 specifically refers to:

[0021] Based on the roulette wheel method, we combine wDestroy and wRepair, select one Destroy operator and one Repair operator to form a combination of Destroy and Repair operators to perform a neighborhood search on the initial cutting order, obtain a new cutting order, and update the initial cutting order based on the new cutting order.

[0022] The Metropolice criterion is used to update the sequential code and score the new sequential code. If the score is greater than the score of the global optimal solution, let W = W1. If the score is greater than the score of the current optimal solution, let W = W2. If the score is less than or equal to the score of the current optimal solution, the new cutting order is accepted with probability p. If accepted, let W = W3; otherwise, let W = W4, where W1 > W2 > W3 > W4; W represents the score of the operator.

[0023] The update formulas for destroyScore and repairScore are:

[0024] destroyScore[destroyOperatorIndex]+=W;

[0025] repairScore[repairOperatorIndex]+=W;

[0026] In the formula: destroyOperatorIndex is the index of the Destroy operator, and repairOperatorIndex is the index of the Repair operator.

[0027] Optionally, in S14, the update formulas for wDestroy and wRepair are:

[0028] wDestroy[destroyOperatorIndex]=wDestroy[destroyOperatorIndex]*b+

[0029] (1-b)*(destroyScore[destroyOperatorIndex] / destroyUseTimes[destroyOperatorIndex]);

[0030] wRepair[repairOperatorIndex]=wRepair[repairOperatorIndex]*b+

[0031] (1-b)*(repairScore[repairOperatorIndex] / repairUseTimes[repairOperatorIndex]);

[0032] In the formula: destroyUseTimes is the number of times the Destroy operator is executed, repairUseTimes is the number of times the Re pair operator is executed, and b is a constant coefficient.

[0033] Optionally, in S22, X = [x1, x2, ..., x n [v1, v2, ..., v] represents the optimal cutting sequence for the steel plate, with the combined code [v1, v2, ..., v] n The symbol indicates that the cutting path starts from the origin and begins cutting steel plate x1 from vertex v1. After that, it continues cutting part x2, starting from vertex v2 of part x2, and so on, until the cutting path finally starts from vertex v2 of part x2. n v n The vertex returns to the origin to complete the cutting process.

[0034] Optionally, S23 specifically involves: for each initial set of cutting starting points, randomly selecting several cutting starting points and replacing them with the numbers of the remaining vertices of the same steel plate to obtain a new set of cutting starting points.

[0035] This invention also provides a laser cutting path planning system for steel plates based on step-by-step optimization, comprising:

[0036] The steel plate cutting sequence module is used to determine the optimal cutting sequence for steel plates.

[0037] The numbered distance unit is used to number each steel plate in the steel plate layout result diagram, abstract each steel plate as a point, and then calculate the distance matrix between all steel plates;

[0038] The initialization unit is used to generate an initial cutting order randomly by using an adaptive large-range neighborhood search simulated annealing algorithm based on the distance matrix, and to encode the numbers of the steel plates according to the initial cutting order. It also initializes the weights wDestroy and wRepair of the neighborhood search operator, and initializes the scores destroyScore and repairScore of the neighborhood search operator.

[0039] The sequential update unit is used to select a search operator based on wDestroy and wRepair using a roulette wheel approach, generate a new cutting order, update the sequential code using the Metropolice criterion, and score the new sequential code to update the destroyScore and repairScore.

[0040] The weight update unit is used to update the weights wDestroy and wRepair of the neighborhood search operator based on the updated destroyScore and repairScore, and to update the annealing temperature and the current optimal cutting order.

[0041] The first repeating unit is used to repeatedly execute the sequence update unit to the weight update unit until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate.

[0042] The steel plate cutting start point module is used to determine the optimal cutting start point for the steel plate:

[0043] The numbering unit is used to group all the vertices of each steel plate as a set and number each vertex in the set;

[0044] The combined coding unit is used to select the number of one vertex in each steel plate according to the optimal cutting order of the steel plate using a genetic algorithm and combine them for coding. This is recorded as the initial cutting start set, and multiple initial cutting start sets are randomly generated.

[0045] The starting point update unit is used to generate a new cutting starting point set for each initial cutting starting point set using the exchange operator, and to update all initial cutting starting point sets using the Metropolice criterion.

[0046] The second repeating unit is used to repeatedly execute the starting point update unit until the set number of vertex iterations is reached to obtain the optimal cutting starting point set, that is, the optimal cutting starting point for each steel plate.

[0047] The cutting path module is used to obtain the cutting path based on the optimal cutting sequence of the steel plates and the optimal cutting start point of each steel plate.

[0048] The present invention also provides an electronic device comprising: a processor and a memory, wherein the memory stores at least one computer instruction, the instruction being loaded by the processor and executing the steps performed in the above-described step-by-step optimization-based laser cutting path planning method for steel plates.

[0049] The present invention also provides a computer-readable storage medium storing at least one computer instruction, which is loaded by a processor and executed by performing the steps in the above-described step-by-step optimization-based laser cutting path planning method for steel plates; the computer-readable storage medium includes a RAM memory, a magnetic drive, an optical drive, and a floppy disk.

[0050] The effects of this invention are as follows:

[0051] This invention presents a step-by-step optimization-based laser cutting path planning method for steel plates. First, an adaptive large-range neighborhood search simulated annealing algorithm is used to obtain the optimal cutting sequence of the steel plates, minimizing the idle travel of the laser head between parts. Then, a genetic algorithm is used to obtain the optimal cutting starting point for each steel plate. Finally, the cutting path is obtained based on the optimal cutting starting points of the steel plates and the optimal cutting starting points of each steel plate. This step-by-step optimization method solves the path planning problem for large-scale material cutting, resulting in high solution efficiency and helping to reduce cutting time, shorten production cycles, and improve automation levels. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the laser cutting path planning method for steel plates based on step-by-step optimization according to Embodiment 1 of the present invention;

[0053] Figure 2 This is a flowchart of the laser cutting path planning method for steel plates based on step-by-step optimization, according to Embodiment 1 of the present invention.

[0054] Figure 3 This is a schematic diagram of the optimal cutting sequence for the steel plate component in Embodiment 5 of the present invention;

[0055] Figure 4 This is a schematic diagram of the optimal cutting starting point for each steel plate component in Embodiment 5 of the present invention;

[0056] Figure 5 This is a schematic diagram of the cutting path in Embodiment 5 of the present invention. Detailed Implementation

[0057] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

[0058] The specific laser cutting process includes the following steps: First, with the laser output off, the laser head moves from the origin to the cutting start point of the first steel plate. Then, the laser output is turned on to cut the first steel plate. After cutting, the laser output is turned off, and the laser head moves to the cutting start point of the next steel plate to begin cutting. This process is repeated until all steel plates are cut, and finally, the laser head returns to the origin. During the cutting process, the laser head's travel distance when the laser output is off is considered an idle travel distance.

[0059] Generally, a steel plate component includes an inner contour and an outer contour. The contour is a closed figure composed of straight lines and arcs, and the intersection points of straight lines with straight lines, straight lines with arcs, or straight lines with arcs are the vertices of the steel plate component. Theoretically, the starting point for cutting the contour can be any position on the contour, but in practice, the starting point is generally defined as the vertex of the contour. In this invention, the starting point is selected from the vertices of the outer contour, but the final cutting order of the steel plate component is still to cut the inner contour first and then the outer contour. The cutting order inside the steel plate component is determined by the vertex information of the steel plate component. When there are multiple inner contours inside the steel plate component, the internal contour cutting path planning problem is consistent with the multiple outer contour path planning problem of steel plate components. Therefore, the steel plate component cutting path problem can be simplified into a steel plate component outer contour path planning problem.

[0060] Specifically, taking the outer contour path planning as an example, the total distance includes the idle travel and the perimeter of the outer contour of the steel plate. Assuming there are N steel plates to be cut in the layout, and the outer contour of the steel plate r has V... r There are vertices, define dist(v) as a vertex i ,v j ) is the vertex v i and v j The distance between them, if the cutting order of the parts is X, the total distance can be calculated using the following formula:

[0061]

[0062] Among them, v o v is the origin of the coordinate system. X[i] Let X[i] be the starting point for cutting the X[i]th steel plate, and L be the perimeter of the outer contour of all steel plates.

[0063] The mathematical model is established as follows:

[0064]

[0065] For large-scale material cutting problems, the number of steel plate parts is large, and the number of vertices in each part is uncertain, making it difficult to find an exact solution using the model. Solving the model based on intelligent algorithms can obtain a better solution within an acceptable time. If a holistic encoding approach is used, simultaneously searching for the cutting sequence and cutting start point, the model's solution space becomes too large, and the solution process includes a large number of invalid searches. This easily leads to long solution times and poor solution quality, which is not conducive to reducing time and production costs.

[0066] Example 1

[0067] Figure 1 This is a schematic diagram of the laser cutting path planning method for steel plates based on step-by-step optimization according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of the laser cutting path planning method for steel plates based on step-by-step optimization, according to Embodiment 1 of the present invention. Figure 1 and Figure 2 As shown, this invention provides a method for laser cutting path planning of steel plates based on step-by-step optimization, which includes:

[0068] S1, Determine the optimal cutting sequence for the steel plate:

[0069] S11: Number each steel plate in the steel plate layout result diagram, abstract each steel plate as a point, and then calculate the distance matrix between all steel plates. Each steel plate is abstracted as a point; this point can be an actual point, such as the centroid of the steel plate, or an abstract point. The ultimate goal is to calculate the distance matrix between the steel plates, which can be represented by the distance between the centroids or by the shortest distance between the vertices of the outer contours of two polygons.

[0070] S12, based on the distance matrix, an adaptive large-range neighborhood search simulated annealing algorithm is adopted. An initial cutting order is randomly generated, and the steel plate numbers are encoded according to this order. The weights wDestroy and wRepair of the search operators are initialized, as are the scores destroyScore and repairScore. In this embodiment, the neighborhood search of the adaptive large-range neighborhood search simulated annealing algorithm is achieved through a combination of the Destroy and Repair operators. The Destroy operator includes randomDestroy and maxDestroy; the Repair operator includes greedyinsert and randominsert. The initial weights of the Destroy and Repair operators are set to the same, and the scores destroyScore and repairScore are initialized to 0. Figure 2 In this context, T0 represents the annealing temperature setting.

[0071] S13 uses a roulette wheel approach, based on wDestroy and wRepair, to select a search operator, generate a new cutting order, update the order code using the Metropolice criterion, and score the new order code to update destroyScore and repairScore.

[0072] Based on the roulette wheel method, we combine wDestroy and wRepair, select one Destroy operator and one Repair operator to form a combination of Destroy and Repair operators to perform a neighborhood search on the initial cutting order, obtain a new cutting order, and update the initial cutting order based on the new cutting order.

[0073] The Metropolice criterion is used to update the sequential code and score the new sequential code. If the score is greater than the score of the global optimal solution, then let W = W1. If the score is greater than the score of the current optimal solution but not higher than the score of the global optimal solution, then let W = W2. If the score is less than or equal to the score of the current optimal solution, then the new cutting order is accepted with probability p. If accepted, then let W = W3; otherwise, let W = W4, where W1 > W2 > W3 > W4; W represents the score of the operator.

[0074] The update formulas for destroyScore and repairScore are:

[0075] destroyScore[destroyOperatorIndex]+=W;

[0076] repairScore[repairOperatorIndex]+=W;

[0077] In the formula: destroyOperatorIndex is the index of the Destroy operator, and repairOperatorIndex is the index of the Repair operator.

[0078] The combinations of the Destroy and Repair operators are shown in Table 1.

[0079] Table 1. Combination methods of the Destroy and Repair operators

[0080]

[0081] S14, update the weights wDe stroy and wRepair of the search operator based on the updated destroyScore and repairScore, and update the annealing temperature and the current optimal cutting order.

[0082] The update formulas for wDestroy and wRepair are:

[0083] wDestroy[destroyOperatorIndex]=wDestroy[destroyOperatorIndex]*b+

[0084] (1-b)*(destroyScore[destroyOperatorIndex] / destroyUseTimes[destroyOperatorIndex]);

[0085] wRepair[repairOperatorIndex]=wRepair[repairOperatorIndex]*b+

[0086] (1-b)*(repairScore[repairOperatorIndex] / repairUseTimes[repairOperatorIndex]);

[0087] In the formula: destroyUseTimes is the number of times the Destroy operator is executed, repairUseTimes is the number of times the Re pair operator is executed, and b is a constant coefficient. In this embodiment, b is 0.5.

[0088] S15, repeat S13-S14 until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate.

[0089] S2, Determine the optimal cutting starting point for the steel plate:

[0090] S21, take all the vertices of each steel plate as a set, and number each vertex in the set.

[0091] S22. Using a genetic algorithm, the number of one vertex in each steel plate is selected according to the optimal cutting order of the steel plate and combined and encoded, which is recorded as the initial cutting starting point set. Multiple initial cutting starting point sets are randomly generated.

[0092] Let X = [x1, x2, ..., x n [v1, v2, ..., v] represents the optimal cutting sequence for the steel plate, with the combined code [v1, v2, ..., v] n The symbol indicates that the cutting path starts from the origin and begins cutting steel plate x1 from vertex v1. After that, it continues cutting part x2, starting from vertex v2 of part x2, and so on, until the cutting path finally starts from vertex v2 of part x1. n v n The vertex returns to the origin to complete the cutting process.

[0093] S23, for each initial cutting starting point set, a new cutting starting point set is generated using the exchange operator, and all initial cutting starting point sets are updated using the Metropolice criterion. For each initial cutting starting point set, several cutting starting points are randomly selected and replaced with the numbers of the remaining vertices of the same steel plate to obtain a new cutting starting point set.

[0094] S24. Repeat S23 until the set number of iterations is reached to obtain the optimal set of cutting starting points, that is, the optimal cutting starting point for each steel plate.

[0095] S3, the cutting path is obtained based on the optimal cutting starting point of the steel plate and the optimal cutting starting point of each steel plate.

[0096] The cutting path generates corresponding G-code to separate and cut all the steel plates in the layout result diagram.

[0097] Example 2

[0098] This invention also provides a laser cutting path planning system for steel plates based on step-by-step optimization, comprising:

[0099] The steel plate cutting sequence module is used to determine the optimal cutting sequence for steel plates.

[0100] The numbered distance unit is used to number each steel plate in the steel plate layout result diagram, abstracting each steel plate as a point, and then calculating the distance matrix between all steel plates.

[0101] The initialization unit is used to generate an initial cutting order randomly based on the distance matrix using an adaptive large-range neighborhood search simulated annealing algorithm, and to encode the numbers of the steel plates according to the initial cutting order. It also initializes the weights wDestroy and wRepair of the neighborhood search operator, and initializes the scores destroyScore and repairScore of the neighborhood search operator.

[0102] The sequential update unit is used to select a search operator based on wDestroy and wRepair using a roulette wheel approach, generate a new cutting order, update the sequential code using the Metropolice criterion, and score the new sequential code to update the destroyScore and repairScore.

[0103] The weight update unit is used to update the weights wDestroy and wRepair of the neighborhood search operator based on the updated destroyScore and repairScore, and to update the annealing temperature and the current optimal cutting order.

[0104] The first repeating unit is used to repeatedly execute the sequence update unit to the weight update unit until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate.

[0105] The steel plate cutting start point module is used to determine the optimal cutting start point for the steel plate:

[0106] The numbering unit is used to group all the vertices of each steel plate as a set and number each vertex in the set.

[0107] The combined coding unit is used to select the number of one vertex in each steel plate according to the optimal cutting order of the steel plate using a genetic algorithm, and combine and encode it as the initial cutting starting point set. Multiple initial cutting starting point sets are randomly generated.

[0108] The starting point update unit is used to generate a new cutting starting point set for each initial cutting starting point set using the exchange operator, and to update all initial cutting starting point sets using the Metropolice criterion.

[0109] The second repeating unit is used to repeatedly execute the starting point update unit until the set number of vertex iterations is reached to obtain the optimal cutting starting point set, that is, the optimal cutting starting point for each steel plate.

[0110] The cutting path module is used to obtain the cutting path based on the optimal cutting sequence of the steel plates and the optimal cutting start point of each steel plate.

[0111] Example 3

[0112] The present invention also provides an electronic device comprising: a processor and a memory, wherein the memory stores at least one computer instruction, the instruction being loaded by the processor and executing the steps performed in the above-described step-by-step optimization-based laser cutting path planning method for steel plates.

[0113] Example 4

[0114] The present invention also provides a computer-readable storage medium storing at least one computer instruction, which is loaded by a processor and executed by performing the steps in the above-described step-by-step optimization-based laser cutting path planning method for steel plates; the computer-readable storage medium includes a RAM memory, a magnetic drive, an optical drive, and a floppy disk.

[0115] Example 5

[0116] Specifically, a specific example is given using the method of the present invention:

[0117] The layout result diagram includes 6 steel plate components. The vertex coordinates of the corresponding outer contour (polygon) of each steel plate component are as follows:

[0118] Polygon1=[[0,0],[5,0],[5,2],[0,2]];

[0119] Polygon2=[[6,0],[8,0],[7,5]];

[0120] Polygon3=[[9,0],[11,0],[11,4],[9,3]];

[0121] Polygon4=[[2,3],[5,3],[5,6],[0,6]];

[0122] Polygon5=[[6,6],[7,7],[10,7],[10,10],[6,10]];

[0123] Polygon6=[[0,8],[1,7],[4,7],[4,9],[3,10],[3,8]];

[0124] Polygon represents a polygon.

[0125] The coordinates of each steel plate component, abstracted as a single point, are as follows:

[0126] Polygon 1 = (2.5, 1);

[0127] Polygon 2 = (7, 1.67);

[0128] Polygon 3 = (10.0, 1.762);

[0129] Polygon 4 = (2.96, 4.625);

[0130] Polygon 5 = (7.94, 8.43);

[0131] Polygon 6 = (2.6, 7.9);

[0132] The distance matrix is ​​shown in Table 2.

[0133] Table 26 Distance Matrix Between Steel Plates

[0134]

[0135] The optimal cutting sequence for the 6 steel plates is as follows: Figure 3 As shown, the optimal cutting starting points for the six steel plates are as follows: Figure 4 As shown, the final cutting path is as follows: Figure 5 As shown.

[0136] This invention avoids many inefficient calculations. For example, if the distance between sequentially cut steel plates is too large, there is no need to optimize the cutting start point of the steel plates, which can greatly improve the calculation efficiency. Finally, the cutting path may not be the optimal solution, but considering the solution time and the quality of the solution, step-by-step optimization of the cutting path can save time and cost, improve operating efficiency and the automation level of the production line.

[0137] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Any modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for laser cutting path planning of steel plates based on step-by-step optimization, characterized in that, It includes: S1, Determine the optimal cutting sequence for the steel plate: S11, number each steel plate in the steel plate layout result diagram, obtain the distance between the centroids of different steel plates, and then calculate the distance matrix between all steel plates; S12, based on the distance matrix, adopts an adaptive large-range neighborhood search simulated annealing algorithm, randomly generates the initial cutting order, encodes the numbers of the steel plates according to the initial cutting order, initializes the weights wDestroy and wRepair of the search operator, and initializes the scores destroyScore and repairScore of the search operator; S13 uses a roulette wheel method, based on wDestroy and wRepair, to select a search operator, generate a new cutting order, update the order code using the Metropolice criterion, and score the new order code to update destroyScore and repairScore. S13 specifically refers to: Based on the roulette wheel method, we combine wDestroy and wRepair, select one Destroy operator and one Repair operator to form a combination of Destroy and Repair operators to perform a neighborhood search on the initial cutting order, obtain a new cutting order, and update the initial cutting order based on the new cutting order. The Metropolice criterion is used to update the sequential code and score the new sequential code. If the score is greater than the score of the global optimal solution, then let W=W1. If the score is greater than the score of the current optimal solution, then let W=W2. If the score is less than or equal to the score of the current optimal solution, then the new cutting order is accepted with probability p. If accepted, then let W=W3; otherwise, let W=W4, where W1>W2>W3>W4; W represents the score of the operator. The update formulas for destroyScore and repairScore are: ； ； In the formula: For the index of the Destroy operator, For the index of the Repair operator; S14, update the weights wDestroy and wRepair of the search operator based on the updated destroyScore and repairScore, and update the annealing temperature and the current optimal cutting order; In S14, the update formulas for wDestroy and wRepair are: ； ； In the formula: The number of times the Destroy operator is executed. The number of times the Repair operator is executed, where b is a constant coefficient; S15, repeat S13-S14 until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate; S2, determine the optimal cutting starting point for the steel plate: S21, take all the vertices of each steel plate as a set, and number each vertex in the set; S22. Using a genetic algorithm, the number of one vertex in each steel plate is selected according to the optimal cutting order of the steel plate and combined and encoded, which is recorded as the initial cutting starting point set. Multiple initial cutting starting point sets are randomly generated. In S22, The optimal cutting sequence for the steel plate is coded as follows: This indicates that the cutting path starts from the origin and begins from the steel plate. of Vertex starts on steel plate Perform cutting, and then continue cutting parts. The cutting starting point is the part of Vertex, and so on, eventually from parts of The vertex returns to the origin to complete the cutting process; S23, For each initial cutting starting point set, use the exchange operator to generate a new cutting starting point set, and use the Metropolice criterion to update all initial cutting starting point sets; S24, repeat S23 until the set number of iterations is reached to obtain the optimal cutting starting point set, that is, the optimal cutting starting point for each steel plate. S3, the cutting path is obtained based on the optimal cutting sequence of the steel plates and the optimal cutting starting point of each steel plate.

2. The method for laser cutting path planning of steel plates based on step-by-step optimization according to claim 1, characterized in that, The neighborhood search of the simulated annealing algorithm for adaptive large-range neighborhood search is achieved by combining the Destroy and Repair operators. The Destroy operator includes randomDestroy and maxDestroy; the Repair operator includes greedyinsert and randominsert. The initial weights of the Destroy and Repair operators are set to be the same, and the scores of the Destroy and Repair operators, destructionScore and repairScore, are initialized to 0.

3. The method for laser cutting path planning of steel plates based on step-by-step optimization according to claim 1, characterized in that, S23 specifically involves: for each initial set of cutting starting points, randomly selecting several cutting starting points and replacing them with the numbers of the remaining vertices of the same steel plate to obtain a new set of cutting starting points.

4. A step-by-step optimization-based laser cutting path planning system for steel plates, the system being used to implement the step-by-step optimization-based laser cutting path planning method for steel plates as described in claim 1, characterized in that, The system includes: The steel plate cutting sequence module is used to determine the optimal cutting sequence for steel plates. The numbered distance unit is used to number each steel plate in the steel plate layout result diagram, obtain the distance between the centroids of different steel plates, and then calculate the distance matrix between all steel plates; The initialization unit is used to generate an initial cutting order randomly by using an adaptive large-range neighborhood search simulated annealing algorithm based on the distance matrix, and to encode the numbers of the steel plates according to the initial cutting order. It also initializes the weights wDestroy and wRepair of the neighborhood search operator, and initializes the scores destroyScore and repairScore of the neighborhood search operator. The sequential update unit is used to select a search operator based on wDestroy and wRepair using a roulette wheel approach, generate a new cutting order, update the sequential code using the Metropolice criterion, and score the new sequential code to update the destroyScore and repairScore. The weight update unit is used to update the weights wDestroy and wRepair of the neighborhood search operator based on the updated destroyScore and repairScore, and to update the annealing temperature and the current optimal cutting order. The first repeating unit is used to repeatedly execute the sequence update unit to the weight update unit until the set number of iterations is reached to obtain the optimal cutting sequence of the steel plate. The steel plate cutting start point module is used to determine the optimal cutting start point for the steel plate: The numbering unit is used to group all the vertices of each steel plate as a set and number each vertex in the set; The combined coding unit is used to select the number of one vertex in each steel plate according to the optimal cutting order of the steel plate using a genetic algorithm and combine them for coding. This is recorded as the initial cutting start set, and multiple initial cutting start sets are randomly generated. The starting point update unit is used to generate a new cutting starting point set for each initial cutting starting point set using the exchange operator, and to update all initial cutting starting point sets using the Metropolice criterion. The second repeating unit is used to repeatedly execute the starting point update unit until the set number of iterations is reached to obtain the optimal cutting starting point set, that is, the optimal cutting starting point for each steel plate. The cutting path module is used to obtain the cutting path based on the optimal cutting sequence of the steel plates and the optimal cutting start point of each steel plate.

5. An electronic device, characterized in that, It includes: The processor and memory, wherein the memory stores at least one computer instruction, which is loaded by the processor and executes the steps performed in the step-by-step optimization-based laser cutting path planning method for steel plates as described in any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The storage medium stores at least one computer instruction, which is loaded by a processor and executed as the steps performed in the step-by-step optimization-based laser cutting path planning method for steel plates according to any one of claims 1-3; the computer-readable storage medium includes RAM memory, magnetic drive, optical drive and floppy disk.

Images