An optimization method for cutting tube profiles considering chamfering

Abstract

A method for optimizing the cutting of pipe profiles with bevel cutting in mind includes the following steps: S1. Generates map data for the row spacing of all beveled parts to facilitate rapid data acquisition for subsequent layout; S2. Generates all feasible cutting columns for the maximum-length pipe profile material to ensure a universally feasible solution; S3. Screens feasible cutting columns that compress length in the case of bevel cutting, and simultaneously determines feasibility after extending the material length for bevel cutting; S4. Uses a Cplex-based integer programming model to solve the one-dimensional stock cutting problem for the obtained feasible columns; S5. Further optimizes large-scale problems in actual production processes using a solution space search scheme based on grouped iterative optimization. Different strategies for invoking the integer programming model are used to solve problems of different scales. This solution uses a specific optimization strategy to group the pipe profiles to obtain base solutions and combines them, narrowing the solution space of a single integer programming, thereby significantly reducing the overall problem-solving time.

Description

Technical Field

[0001] The invention relates to a pipe cutting strategy, in particular to a pipe profile cutting optimization method considering beveling. Background Art

[0002] The one-dimensional cutting stock problem (1D CSP) is a classic combinatorial optimization problem in production and operations management, widely used in manufacturing, wood processing, metal cutting, and textile industries. The goal of this problem is to arrange items of various required lengths given a given length of raw material, so as to minimize the amount of raw material used or maximize the utilization rate. In actual production, factories often need to cut parts of different lengths from raw materials. For example, in the parts manufacturing industry, pipe profiles of various lengths are cut from tube blanks. However, direct cutting may generate a large amount of waste, which not only wastes resources but also increases production costs. Therefore, how to reasonably arrange the cutting sequence and plan to minimize the amount of waste is an important issue for enterprises to optimize production.

[0003] For traditional pipe fitting manufacturers, the production demand for pipe fittings is generally based on straight cutting. The optimization method for straight cutting of pipe profiles is relatively complete and mature, and classic one-dimensional cutting optimization algorithms such as greedy algorithm, dynamic programming, and column generation are often used. However, as industry demand gradually diversifies, for the different uses of pipe profiles and user personalized needs, traditional pipe profile straight cutting layout algorithms do not consider the optimization of cross-sections at different angles, and cannot meet customers' diverse usage and installation needs. The existing technology performs poorly in terms of pipe profile utilization, resulting in a large amount of material being wasted. In addition, a single pipe profile needs to be cut many times and a large amount of beveled leftover material cannot be reused for secondary nesting, further affecting production efficiency and pipe profile utilization.

[0004] While combining all possible permutations might be feasible for a small number of tubing components, exhaustively enumerating all possible layout options for larger quantities of tubing and complex profiles would result in an extremely large number of possible solutions. One-dimensional layout optimization is an NP-hard problem. Directly using integer programming to solve the vast search and solution space for a layout problem would result in a lengthy computational process. This approach cannot guarantee convergence, ultimately rendering the entire solution process extremely lengthy. Summary of the Invention

[0005] To address the above defects, a solution space search scheme based on group optimization is proposed. It generates a large number of layout schemes as the search space through a greedy strategy, and groups the pipe parts through a certain optimization strategy to obtain basic solutions for combination, thereby narrowing the solution space of single integer programming and greatly reducing the time to solve the entire problem.

[0006] To achieve this object, the present invention adopts the following technical solutions:

[0007] A method for optimizing pipe profile cutting considering beveling, comprising the following steps:

[0008] S1. Generate Map data of the row spacing of all beveled parts to facilitate the rapid acquisition of subsequent nesting data;

[0009] S2. Generate all feasible cutting sequences for the maximum length tube profile material to ensure a universal feasible solution;

[0010] S3. Screen the feasible cutting columns that consider length compression in the case of bevel cutting, and make a feasibility judgment on the length extension of the raw material in the case of bevel cutting;

[0011] S4. Solve the one-dimensional cutting stock problem based on the obtained feasible sequence using the Cplex integer programming model;

[0012] S5. For large-scale problems in actual production processes, a solution space search scheme based on group iterative optimization is used for further optimization. Different strategies are used to call the integer programming model to solve problems of different scales.

[0013] Preferably, in step S1, for a large number of parts containing beveled edge data, it is necessary to obtain the distance between any two parts when placed in any order and position, so as to facilitate the search for the optimal layout solution. The data of the compressible distance matrix is obtained by performing collision detection after unfolding the parts. The standard lengths of a group of parts containing beveled edges are L1 and L2, the compressible distance is ΔL, and the combined length is L1+L2-ΔL;

[0014] The data of all beveled parts are processed and summarized into a part spacing data map.

[0015] Preferably, for the generation of the initial feasible solution of the tube profile cutting plan, a method is proposed to recursively and quickly generate all feasible cutting methods using the greedy idea, that is, first sort by the cutting length, and within the range of the raw material length L, first arrange the long material blanks to generate the first cutting method, and then take off a long material in the first cutting method, arrange a medium-length material, and then place the remaining material with shorter cuttings to form a new cutting method; each cutting method is formed on the basis of the latest generated cutting method, and only a small number of cuttings need to be changed to generate another new cutting method, which reduces a lot of unnecessary calculations and can quickly generate all feasible methods; in each cutting method, the number A of the nth type of parts n Should not be greater than the demand for the part b n , where the blanks are arranged in descending order of length from l1 to l n, the current remaining length of the i-th cutting material in the process of generating the cutting method is S i (i=1,2,3,…,n) represents.

[0016] Preferably, in step S3, since the tubular members to be arranged may contain one or two beveled edges, if the arrangement is performed based on the maximum axial distance of the tubular members, the length of the subsequent single tubular members will inevitably be shortened to a certain extent. Therefore, when the tubular members with beveled edges are equivalently replaced with straight-cut tubular members, the length of the tubular material needs to be increased by a certain length L′, thereby expanding the solution space range.

[0017] Feasibility assessment of bevel cutting layout: The series of layout solutions obtained by the greedy strategy recursive method mentioned above may include the optimal layout solution for the original problem, or may not meet the requirements of the raw material length no matter how they are sorted. It is necessary to judge the feasibility of the solutions to select the layout solutions that meet the constraints of the original problem. Therefore, a reasonable method is needed to evaluate the feasibility of each layout optimization solution.

[0018] The actual length of the raw material pipe profile is used as the evaluation indicator. That is, the raw material utilization rate of each solution is calculated through a greedy algorithm. The solution that meets the length constraint and has the highest utilization rate is selected. The purpose is to maximize the screening and reduce the cost of single raw material layout and minimize the total cost of the cutting process. The process of judging the feasibility of beveling for the initial layout solution is as follows:

[0019] Step 1: Initialize the plan information in the nesting set, select a nesting plan, and obtain the types and quantities of pipe parts in the plan;

[0020] Step 2: Randomly select a pipe with a length of l0 from the plan for initial placement;

[0021] Step 3: Obtain the tube-shaped part l with the largest compressible distance to the left of the tube-shaped part l0 in the solution from the map a And the tubular part with the largest compressible distance on the right side of the tubular part l b , directly obtain the total compressible distance Δl through map; for the actual oblique cutting length l m Perform calculations;

[0022] l m =l0+l a +l b -Δl(3.1)

[0023] When l m >L, skip to step 2 and reselect the pipe parts for initial placement;

[0024] When l m <=L, jump to step 4;

[0025] Step 4: Update the length of l0 so that the arrangement of the tubular components can be considered as a whole for subsequent tubular component additions;

[0026] Step 5: Determine whether all the tubes in the solution have been arranged. If not, jump to step 4. If they have been arranged, retain the optimal solution, that is, the solution with the shortest total arrangement length.

[0027] Step 6: Determine whether the solution set has been sorted. If not, jump to step 2 and select a sorting solution. If it has been sorted, the algorithm ends.

[0028] Preferably, in step S4: given a single specification long strip raw material tube profile, the length of which is L, there are n types of oblique cylindrical tube parts of different specifications to be cut, and the length of each workpiece is l i , the number is d i , i∈[1,n], it is required to arrange all types and quantities of tube parts on the tube profile without overlapping, so that the amount of tube profile consumed in cutting is the least;

[0029] Assume that the feasible arrangement method set P = [p1, p2, ..., p g ] contains g kinds of arrangement methods, and the number of times the jth arrangement method is used is x j The number of tube-shaped parts of type i in this arrangement is a ij (i=1,2,…,n);

[0030] The integer programming model with the minimum number of tube parts as the optimization goal is as follows:

[0031] The solution goal is:

[0032] The constraints are:

[0033] For this type of integer programming problem, this paper uses the Cplex solver to quickly find the optimal solution to the target value.

[0034] Furthermore, in step S5: since the number of solutions generated by the recursive algorithm of the greedy strategy is uncertain, in integer programming, one variable in the objective function corresponds to one packing solution and one feasible column in the constraint condition. For a solution space with a small column size, integer programming can be directly used for solving to ensure the quality and speed of the solution. However, for problems with a large column size and hundreds of thousands or even millions of variables in the objective function, directly using integer programming will result in an extremely long solution time and cannot guarantee the solution quality. Therefore, it is necessary to quickly and accurately find the base column in the huge solution space and eliminate useless packing solutions. The solution space search scheme based on group optimization in this method is as follows:

[0035] First, the size n of the layout plan generated by the problem is judged. If the solution size n is less than or equal to 10,000, the pipe profile layout integer programming is directly used for solution. The total number of raw materials used and the number of times each solution is used can be quickly obtained.

[0036] If the solution size is greater than n>10000, the group optimization search method is used to divide the solutions into m groups with equal spacing, and the solution S in group i is i The number of feasible sequences in is s i (i=1,2,...,m), satisfying The large-scale solution space is divided into multiple small-scale solution spaces for solution, but for each solution space, the integer

[0037] The solution obtained by planning is only a local optimal solution. If you want to solve the global optimal solution, you need to combine the basis columns of each solution space;

[0038] For integer programming of S1, we can solve its basis solution P1, and the corresponding basis column is R1. Then for integer programming of S2, we need to expand the column size and let S2 = {S2, R1} to ensure that the solution of S2 fully considers the optimal solution of S1. Similarly, for S i Perform integer programming and let S i ={S i ,R1,R2,...,R i-1}, we can solve for Gilead as R i , if i = n, the solution of the basic solution P n That is the global optimal solution, Gilead R n This is the global optimal solution.

[0039] One of the above technical solutions includes the following beneficial effects:. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic diagram of an example of a bevel-cut tube as described in the background of the present invention;

[0041] Figure 2 This is the overall algorithm implementation idea of an embodiment of the present invention;

[0042] Figure 3 FIG. 1 is a schematic diagram of the unfolded parts of an embodiment of the present invention, showing how the compressible distance ΔL is generated;

[0043] Figure 4 is a flowchart of greedy recursive feasible column generation according to an embodiment of the present invention;

[0044] Figure 5 is a flow chart of a bevel cutting feasibility determination according to an embodiment of the present invention;

[0045] Figure 6 This is a diagram of a pipe cutting optimization system - straight cutting case according to an embodiment of the present invention;

[0046] Figure 7 This is a pipe cutting optimization system - bevel cutting case diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following describes embodiments of the present invention in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present invention and are not to be construed as limiting the present invention.

[0048] like Figure 2 As shown, a pipe profile cutting optimization method considering bevel cutting includes the following steps:

[0049] S1. Generate Map data of the row spacing of all beveled parts to facilitate the rapid acquisition of subsequent nesting data;

[0050] S2. Generate all feasible cutting sequences for the maximum length tube profile material to ensure a universal feasible solution;

[0051] S3. Screen the feasible cutting columns that consider length compression in the case of bevel cutting, and make a feasibility judgment on the length extension of the raw material in the case of bevel cutting;

[0052] S4. Solve the one-dimensional cutting stock problem based on the obtained feasible sequence using the Cplex integer programming model;

[0053] S5. For large-scale problems in actual production processes, a solution space search scheme based on group iterative optimization is used for further optimization. Different strategies are used to call the integer programming model to solve problems of different scales.

[0054] The present invention uses a pipe bevel optimization algorithm based on equivalent substitution, which regards pipe parts containing beveled edges as "compressible" pipe parts. The length of such pipe parts is not a fixed value, and its length will vary with its arrangement position on the pipe profile, but will not exceed its axial maximum value. That is, arranging specific pipe parts in a certain order may "compress" a part of the length, and the compressed length will not exceed the sum of the axial distances of all its beveled edges. Therefore, a certain extension length can be added to the length of the original pipe profile as compressible space, and a single bevel layout algorithm is called to screen the layout scheme that meets the compression conditions. The bevel problem is converted into a straight cutting problem for integer programming solution, which reduces the complexity of the algorithm, improves the optimization efficiency of the algorithm, and ensures the quality of the solution.

[0055] like Figure 3As shown, in step S1, for a large number of parts containing beveled edge data, it is necessary to obtain the distance between any two parts when placed in any order and position, so as to search for the optimal layout solution. The data of the compressible distance matrix is obtained by performing collision detection after unfolding the parts. The standard lengths of a group of parts containing beveled edges are L1 and L2, the compressible distance is ΔL, and the combined length is L1+L2-ΔL;

[0056] The data of all beveled parts are processed and summarized into a part spacing data map.

[0057] In this way, the bevel features of the parts are converted into compressible specific length values and corresponded through the distance matrix. The data query is simple and intuitive, which greatly simplifies the distance calculation of subsequent layout.

[0058] like Figure 4 As shown, for the generation of the initial feasible solution of the tube profile cutting plan, a method is proposed to recursively and quickly generate all feasible cutting methods using the greedy idea, that is, first sort by the cutting length, and within the range of the raw material length L, first arrange the long material blanks to generate the first cutting method, and then take off a long material in the first cutting method, arrange a medium-length material, and then place the remaining material with shorter cuttings to form a new cutting method; each cutting method is formed on the basis of the latest generated cutting method, and only a small number of cuttings needs to be changed to generate another new cutting method, which reduces a lot of unnecessary calculations and can quickly generate all feasible methods; in each cutting method, the number A of the nth type of parts n Should not be greater than the demand for the part b n , where the blanks are arranged in descending order of length from l1 to l n , the current remaining length of the i-th cutting material in the process of generating the cutting method is S i (i=1,2,3,…,n) represents.

[0059] A series of initial solutions are quickly generated through a greedy strategy, which reduces the complexity of algorithm design while ensuring the quality of the generated solutions.

[0060] like Figure 5 As shown, in step S3, since the tubular parts to be arranged may contain one or two beveled edges, if the tubular parts are arranged according to the maximum axial distance, the length of the subsequent single tubular profiles will inevitably be shortened to a certain extent. Therefore, when the tubular parts with beveled edges are equivalently replaced with straight-cut tubular parts, the length of the tubular profile raw material needs to be increased by a certain length L′, thereby expanding the solution space range;

[0061] Feasibility assessment of bevel cutting layout: The series of layout solutions obtained by the greedy strategy recursive method mentioned above may include the optimal layout solution for the original problem, or may not meet the requirements of the raw material length no matter how they are sorted. It is necessary to judge the feasibility of the solutions to select the layout solutions that meet the constraints of the original problem. Therefore, a reasonable method is needed to evaluate the feasibility of each layout optimization solution.

[0062] The actual length of the raw material pipe profile is used as the evaluation indicator. That is, the raw material utilization rate of each solution is calculated through a greedy algorithm. The solution that meets the length constraint and has the highest utilization rate is selected. The purpose is to maximize the screening and reduce the cost of single raw material layout and minimize the total cost of the cutting process. The process of judging the feasibility of beveling for the initial layout solution is as follows:

[0063] Step 1: Initialize the plan information in the nesting set, select a nesting plan, and obtain the types and quantities of pipe parts in the plan;

[0064] Step 2: Randomly select a pipe with a length of l0 from the plan for initial placement;

[0065] Step 3: Obtain the tube-shaped part l with the largest compressible distance to the left of the tube-shaped part l0 in the solution from the map a And the tubular part with the largest compressible distance on the right side of the tubular part l b , directly obtain the total compressible distance Δl through map; for the actual oblique cutting length l m Perform calculations;

[0066] l m =l0+l a +l b -Δl(3.1)

[0067] When l m >L, skip to step 2 and reselect the pipe parts for initial placement;

[0068] When l m <=L, jump to step 4;

[0069] Step 4: Update the length of l0 so that the arrangement of the tubular components can be considered as a whole for subsequent tubular component additions;

[0070] Step 5: Determine whether all the tubes in the solution have been arranged. If not, jump to step 4. If they have been arranged, retain the optimal solution, that is, the solution with the shortest total arrangement length.

[0071] Step 6: Determine whether the solution set has been sorted. If not, jump to step 2 and select a sorting solution. If it has been sorted, the algorithm ends.

[0072] The validity of the nesting scheme is verified by a heuristic algorithm. By finding the parts with the largest compressible distance in the scheme from both ends, it is possible to quickly verify whether the generated scheme meets the length constraint of the bevel nesting, thereby ensuring the feasibility of the scheme nesting.

[0073] In addition, in step S4: given a single specification long strip raw material tube profile, the length of which is L, there are n types of oblique cylindrical tube parts of different specifications to be cut, and the length of each workpiece is l i , the number is d i , i∈[1,n], it is required to arrange all types and quantities of tube parts on the tube profile without overlapping, so that the amount of tube profile consumed in cutting is the least;

[0074] Assume that the feasible arrangement method set P = [p1, p2, ..., p g ] contains g kinds of arrangement methods, and the number of times the jth arrangement method is used is x j The number of tube-shaped parts of type i in this arrangement is a ij (i=1,2,…,n);

[0075] The integer programming model with the minimum number of tube parts as the optimization goal is as follows:

[0076] The solution goal is:

[0077] The constraints are:

[0078] For this type of integer programming problem, this paper uses the Cplex solver to quickly find the optimal solution to the target value.

[0079] The integer programming problem is solved accurately by the solver, ensuring the quality of the solution.

[0080] In addition, in step S5: since the number of solutions generated by the recursive algorithm of the greedy strategy is uncertain, in integer programming, one variable in the objective function corresponds to one packing solution and one feasible column in the constraint condition. For a solution space with a small column size, integer programming can be directly used to solve it, ensuring the quality and speed of the solution. However, for problems with a large column size and hundreds of thousands or even millions of variables in the objective function, directly using integer programming will result in an extremely long solution time and cannot guarantee the solution quality. Therefore, it is necessary to quickly and accurately find the base column in the huge solution space and eliminate useless packing solutions. The solution space search scheme based on group optimization in this method is as follows:

[0081] First, the size n of the layout scheme (feasible series) generated by the problem is judged. If the solution size n is less than or equal to 10,000, the pipe profile layout integer programming is directly used for solving the problem. The total number of raw materials used and the number of times each solution is used can be quickly obtained.

[0082] If the solution size is greater than n>10000, the group optimization search method is used to divide the solutions into m groups with equal spacing, and the solution S in group i is i The number of feasible sequences in is s i (i=1,2,...,m), satisfying The large-scale solution space is divided into multiple small-scale solution spaces for solution, but for each solution space, the integer

[0083] The solution obtained by planning is only a local optimal solution. If you want to solve the global optimal solution, you need to combine the basis columns of each solution space;

[0084] For integer programming of S1, we can solve its basis solution P1, and the corresponding basis column is R1. Then for integer programming of S2, we need to expand the column size and let S2 = {S2, R1} to ensure that the solution of S2 fully considers the optimal solution of S1. Similarly, for S i Perform integer programming and let S i ={S i ,R1,R2,...,R i-1}, we can solve for Gilead as R i , if i = n, the solution of the basic solution P n That is the global optimal solution, Gilead R n This is the global optimal solution.

[0085] Problems of different scales are classified and discussed, and different solution strategies are used to accelerate the algorithm. The quality of the solution is guaranteed for small-scale problems, while fast solution is guaranteed for large-scale problems, which meets the actual production needs of enterprises.

[0086] Computational experiments and analysis

[0087] 1) Computational experiments

[0088] To verify the feasibility and effectiveness of the equivalent substitution-based tube beveling optimization algorithm and the group-optimized solution space search scheme, we conducted iterative optimization and testing using actual production order data from a custom tube cutting company. The iterative optimization algorithm was implemented in Java and compiled with Eclipse 3.7. All experiments were performed on a 3.4 GHz quad-core Intel i7 CPU running Windows 7 SP1 with 8GB of RAM.

[0089] 2) Case Analysis To verify the feasibility and effectiveness of the proposed pipe profile layout optimization algorithm, experiments were conducted on algorithms reported in relevant literature and test cases provided by actual production enterprises. At the same time, to realize and promote the industrial application of the proposed algorithm, a corresponding software system was developed. To verify the optimization effect and feasibility of the algorithm in different application scenarios of straight cutting and bevel cutting, this paper designed two sets of experiments as follows:

[0090] Experiment 1: The test was conducted on 10 sets of straight cutting test data of tube profiles of different lengths provided by tube profile manufacturer A. The parts were 25 sets of square tubes of the same specifications, and the tube profile length ranged from 5000 to 12500.

[0091] Experiment 2: The test is based on some actual production data provided by tube profile manufacturer A in August 2024, which is divided into 10 groups in total. The tube profile lengths and part specifications may be different. The tube profile types include square tubes, rectangular tubes and angle steels.

[0092] 3) Experiment 1 Calculation and Analysis

[0093] Experiment 1 used 10 sets of pipe profile data provided by pipe profile company A for straight cutting tests. The results were compared with those of the current mainstream commercial pipe cutting software at home and abroad. As shown in Table 2, the nesting comparison results show that the overall experimental test results of this algorithm are significantly better than those of the two current mainstream commercial software. Data sets 2, 3, and 4 all used one less pipe profile than the commercial software, and the algorithm run times were 2.831s, 5.025s, and 20.357s, respectively. The other data sets were all on par with the optimized results of the commercial software. The results of the 10 data sets reduced the number of pipe profiles by three, and the algorithm run time was less than 21s. The experimental results of this chapter's algorithm are significantly better than those of the two current commercial software, saving more raw material and achieving higher cutting efficiency.

[0094] Matching tube profile cutting system such as Figure 6 As shown, the cutting arrangement of parts of different specifications and lengths is visualized, including different cutting methods and the number of cutting times. For each cutting method, the dark part is the arranged parts, and the white part is the cutting residue.

[0095] Table 2 Comparison of straight cutting test data

[0096]

[0097] 3) Experiment 2 Calculation and Analysis

[0098] Experiment 2 used 10 sets of pipe profile data provided by pipe profile company A for bevel cutting tests. The pipe profile types included square pipes, rectangular pipes, and angle steels. The calculation results were compared with those of the current mainstream commercial pipe cutting software at home and abroad. As shown in Table 3, the arrangement results in the table show that the total experimental test results of this algorithm differed from the combined optimal results of the two current mainstream commercial software by only one pipe, and the algorithm runtime was less than 21 seconds. Among them, the result of sequence number 8 was one pipe less than the optimal result of the commercial software, and the algorithm runtime was 3.66 seconds. Sequence numbers 5 and 10 both had one pipe more than the comprehensive results of the commercial software, with algorithm runtimes of 2.85 seconds and 1.19 seconds, respectively. The experimental results of the algorithm in this chapter are significantly better than the optimization results of commercial software 1, and the overall difference is one pipe compared to commercial software 2.

[0099] Matching tube profile cutting system such as Figure 7 As shown in the figure, the cutting arrangement of parts of different specifications and lengths in different task batches is visualized, including different cutting methods and number of cutting times. For each cutting method, the dark part is the arranged parts, the white part is the cutting residual material, and the overlapping part between two adjacent parts represents the beveled compression length.

[0100] Table 3 Comparison of oblique cutting test data

[0101]

[0102] The technical principles of the present invention have been described above with reference to specific embodiments. These descriptions are intended solely to illustrate the principles of the present invention and are not to be construed in any way as limiting the scope of protection of the present invention. Based on the explanations herein, those skilled in the art will readily conceive of other specific embodiments of the present invention without inventive effort, and such embodiments will fall within the scope of protection of the present invention.

Claims

1. A pipe profile cutting optimization method considering beveling, characterized in that: The following steps are involved: S1. Generate Map data of the row spacing of all beveled parts to facilitate the rapid acquisition of subsequent nesting data; S2. Generate all feasible cutting sequences for the maximum length tube profile material to ensure a universal feasible solution; S3. Screen the feasible cutting columns that consider length compression in the case of bevel cutting, and make a feasibility judgment on the length extension of the raw material in the case of bevel cutting; S4. Solve the one-dimensional cutting stock problem based on the obtained feasible sequence using the Cplex integer programming model; S5. For large-scale problems in actual production processes, a solution space search scheme based on group iterative optimization is used for further optimization. Different strategies are used to call the integer programming model to solve problems of different scales; In step S3, the actual length of the raw material tube profile is used as an evaluation indicator. That is, the raw material utilization rate of each solution is calculated through a greedy algorithm. The solution that meets the length constraint and has the highest utilization rate is selected. The purpose is to maximize the screening and reduce the cost of single raw material layout and minimize the total cost of the cutting process. The process of judging the feasibility of beveling for the initial layout solution is as follows: Step 1: Initialize the plan information in the nesting set, select a nesting plan, and obtain the types and quantities of pipe parts in the plan; Step 2: Randomly select a pipe with a length of l0 from the plan for initial placement; Step 3: Obtain the tube-shaped part l with the largest compressible distance to the left of the tube-shaped part l0 in the solution from the map a And the tubular part with the largest compressible distance on the right side of the tubular part l b , directly obtain the total compressible distance Δl through map; for the actual oblique cutting length l m Perform calculations; he m =l0+l a +l b -Δl(3.1) When l m >L, skip to step 2 and reselect the pipe parts for initial placement; When l m <=L, jump to step 4; Step 4: Update the length of l0 so that the arrangement of the tubular components can be considered as a whole for subsequent tubular component additions; Step 5: Determine whether all the tubes in the solution have been arranged. If not, jump to step 4. If they have been arranged, retain the optimal solution, that is, the solution with the shortest total arrangement length. Step 6: Determine whether the solution set has been sorted. If not, jump to step 2 and select a sorting solution. If it has been sorted, the algorithm ends.

2. The tube profile cutting optimization method considering beveling according to claim 1, characterized in that: In step S1, for a large number of parts containing beveled edge data, it is necessary to obtain the distance between any two parts when placed in any order and position, so as to facilitate the search for the optimal layout solution. The data of the compressible distance matrix is obtained by unfolding the parts and performing collision detection. The standard lengths of a group of parts containing beveled edges are L1 and L2, the compressible distance is ΔL, and the combined length is L1+L2-ΔL; The data of all beveled parts are processed and summarized into a part spacing data map.

3. The pipe profile cutting optimization method considering beveling according to claim 2, characterized in that: For the generation of the initial feasible solution of the tube profile cutting plan, a method is proposed to recursively and quickly generate all feasible cutting methods using the greedy idea, that is, first sort by the cutting length, and within the range of the raw material length L, first arrange the long material blanks to generate the first cutting method, and then take off a long material in the first cutting method, arrange a medium-length material, and then place the remaining material with shorter cuttings to form a new cutting method; each cutting method is formed on the basis of the latest generated cutting method, and only a small number of cuttings need to be changed to generate another new cutting method, which reduces a lot of unnecessary calculations and can quickly generate all feasible methods; in each cutting method, the number A of the nth type of parts n Should not be greater than the demand for the part b n , where the blanks are arranged in descending order of length from l1 to l n , the current remaining length of the i-th cutting material in the process of generating the cutting method is S i (i=1,2,3,…,n) represents.

4. The pipe profile cutting optimization method considering beveling according to claim 3, characterized in that: In step S4: Given a single specification long strip raw tube profile, the length of which is L, there are n types of oblique cylindrical tube parts of different specifications that need to be cut, and the length of each workpiece is l i , the number is d i , i∈[1,n], it is required to arrange all types and quantities of tube parts on the tube profile without overlapping, so that the amount of tube profile consumed in cutting is the least; Assume that the feasible arrangement method set P = [p1, p2, ..., p g ] contains g kinds of arrangement methods, and the number of times the jth arrangement method is used is x j The number of tube-shaped parts of type i in this arrangement is a ij (i=1,2,…,n); The integer programming model with the minimum number of tube parts as the optimization goal is as follows: The solution goal is: The constraints are: For this type of integer programming problem, using the Cplex solver can quickly find the optimal solution to the target value.

5. The pipe profile cutting optimization method considering beveling according to claim 4, characterized in that: In step S5: Since the number of solutions generated by the recursive algorithm of the greedy strategy mentioned above is uncertain, in integer programming, one variable in the objective function corresponds to one packing solution and one feasible column in the constraints. For solution spaces with smaller columns, integer programming can be used directly to ensure solution quality and speed. However, for problems with large columns and hundreds of thousands or even millions of variables in the objective function, directly using integer programming will result in an extremely long solution time and cannot guarantee solution quality. Therefore, it is necessary to quickly and accurately find the basis column in the huge solution space and eliminate useless packing solutions. The solution space search scheme based on group optimization is as follows: First, the size n of the layout plan generated by the problem is judged. If the solution size n is less than or equal to 10,000, the pipe profile layout integer programming is directly used for solution. The total number of raw materials used and the number of times each solution is used can be quickly obtained. If the solution size is greater than n>10000, the group optimization search method is used to divide the solutions into m groups with equal spacing, and the solution S in group i is i The number of feasible sequences in is s i (i=1,2,...,m), satisfying The large-scale solution space is divided into multiple small-scale solution spaces for solution, but for each solution space, the integer The solution obtained by planning is only a local optimal solution. If you want to solve the global optimal solution, you need to combine the basis columns of each solution space; For integer programming of S1, we can solve its basis solution P1, and the corresponding basis column is R1. Then for integer programming of S2, we need to expand the column size and let S2 = {S2, R1} to ensure that the solution of S2 fully considers the optimal solution of S1. Similarly, for S i Perform integer programming and let S i ={S i ,R1,R2,...,R i-1 }, we can solve for Gilead as R i , if i = n, the solution of the basic solution P n That is the global optimal solution, Gilead R n This is the global optimal solution.

Images