Warehousing and processing collaborative optimization scheduling method and system for flexible manufacturing unit

Abstract

The invention discloses a storage and processing collaborative optimization scheduling method and system for a flexible manufacturing unit, and relates to the technical field of scheduling in a workshop manufacturing process, and the method comprises the steps: constructing a storage and processing collaborative optimization scheduling model for the flexible intelligent manufacturing unit, the scheduling model considers flexible processing scheduling in the flexible intelligent manufacturing unit and scheduling of goods locations of workpieces in a warehouse at the same time, and the optimization target of the scheduling model is to minimize the maximum completion time; flexible operation related data and machine and storage goods location layout data in the flexible intelligent manufacturing unit are collected, then according to the data, a hybrid optimization algorithm combining a genetic algorithm and variable neighborhood search is adopted to carry out optimization solution on the scheduling model, and a processing and storage scheduling scheme of the flexible intelligent manufacturing unit is obtained. According to the method, the production scheduling and the storage scheduling in the flexible intelligent manufacturing unit containing the storage are optimized, the transportation time can be reasonably optimized, the production efficiency can be improved, the production cycle can be shortened, and great economic benefits are brought.

Description

Technical Field

[0001] The present invention relates to the technical field of scheduling in the workshop manufacturing process, and particularly relates to a method and system for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell. Background Art

[0002] With the development of intelligent manufacturing technology and the improvement of material living standards, traditional mechanical manufacturing is gradually difficult to meet market demands due to its characteristics of large batch, single variety, and low flexibility. A flexible manufacturing cell consists of a machining center, an industrial robot, a numerical control machine tool, a material transportation device, an automated stereoscopic warehouse, etc., and has the flexibility to adapt to the processing of multi-variety products.

[0003] Production scheduling is the basis and key for a flexible manufacturing cell to achieve intelligent manufacturing. The flexible job shop scheduling problem (FJSP) widely exists in the actual production process of manufacturing enterprises. In a flexible manufacturing cell, a numerical control machine tool is the core and an indispensable part, and is often singled out for research.

[0004] In a flexible intelligent manufacturing cell, the warehousing module is a key component. The position of the workpiece directly affects the transportation time of the workpiece from the warehouse to the processing equipment, and thus has an important impact on the processing efficiency of the entire system. At present, the research on automated warehousing systems is relatively rich, but little consideration is given to scheduling optimization together with flexible processing as a complete manufacturing cell.

[0005] In the flexible job shop scheduling problem, transportation time is one of the important factors affecting scheduling optimization. Since workpieces need to flow between machines during different processes, the length of transportation time not only affects the completion time of processing tasks, but also has a chain effect on the idle state of machines, the waiting time of workpieces, and the overall production rhythm. However, in the flexible job shop scheduling problem involving transportation time, existing research usually uses randomly generated transportation times or conducts workshop scheduling simulations based on fixed data sets. Using static / random transportation time estimation leads to distorted path planning, and the optimization of warehousing resources and production scheduling forms independent islands, unable to achieve the overall optimal spatio-temporal efficiency of material flow, thus limiting the accurate description of actual production scenarios and the exploration of optimization potential. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell to solve the problems in the prior art in view of the deficiencies of the above-mentioned prior art.

[0007] The present invention specifically provides the following technical solutions:

[0008] A method for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell includes:

[0009] Collect data related to flexible operations and the layout of machines and storage locations in the flexible intelligent manufacturing cell;

[0010] Consider the flexible machining scheduling in the flexible intelligent manufacturing cell and the scheduling of workpiece storage locations to minimize the makespan as the optimization goal, and construct a scheduling model for the collaborative optimization of storage and machining in the flexible intelligent manufacturing cell;

[0011] Input the data related to flexible operations and the layout of machines and storage locations into the scheduling model to obtain a feasible solution including the processing sequence of operations, the processing machines selected for operations, and the locations of workpieces in the warehouse. Decode the feasible solution to obtain a scheduling plan for the corresponding flexible job shop and the locations of workpieces in the warehouse, and process this scheduling plan through a genetic algorithm. Select the solution with the minimum makespan as the high-quality solution, perform iterative variable neighborhood search on the high-quality solution, and when the number of iterations is satisfied, output the optimal individual in the population. Obtain the scheduling plans for machining and storage in the flexible intelligent manufacturing cell based on the optimal individual in the population.

[0012] Preferably, the obtaining of a feasible solution including the processing sequence of operations, the processing machines selected for operations, and the locations of workpieces in the warehouse is specifically as follows:

[0013] Take the processing sequence of all obtained operations as the first layer, i.e., the operation coding layer; take the processing machines selected for each operation as the second layer, i.e., the machine selection coding layer; take the locations of all workpieces in the warehouse as the third layer, i.e., the location coding layer of workpieces in the warehouse. Represent the feasible solution in the form of the operation coding layer, the machine selection coding layer, and the location coding layer of workpieces in the warehouse, where the specific expression of the location coding Z of workpieces in the warehouse is:

[0014] Z = [Z(h, l)] H×L ;

[0015] where represents the state of this storage unit.

[0016] Preferably, the decoding of the feasible solution to obtain a scheduling plan for the corresponding flexible job shop and the locations of workpieces in the warehouse includes:

[0017] Obtain the processing sequence of all operations according to the operation coding layer, sort them in sequence according to the processing sequence, and obtain the processing machine and the corresponding processing time for this operation according to the machine selection coding layer;

[0018] When processing the first operation of the workpiece, obtain the location of the workpiece in the warehouse according to the location coding layer of workpieces in the warehouse, obtain the transportation time between the warehouse and the machine, and when processing the subsequent operations of the workpiece, calculate the transportation time between different machines;

[0019] Judge the no-load start time of the RGV, obtain the load transportation start time and load transportation end time, and judge the processing start time of this process, calculate the processing end time of this process, and after processing in the order of the process coding layer, obtain the processing end time of all processes with the load transportation start time, load transportation end time and processing start time of all processes, and calculate the makespan C with the load transportation start time and the processing end time of all processes max ; The load is the relationship between the workpiece in the warehouse and the machine and the relationship between the workpiece in different machines.

[0020] Preferably, the specific expression of the makespan is:

[0021]

[0022] where C max is the makespan, is the completion time of process O ij .

[0023] Preferably, obtain the processing time and transportation time of each process of each workpiece in the scheduling model, and the specific expression is:

[0024] TS i,1 = max{TWC i,0 , TC p,q};

[0025]

[0026] TRS i,j-1,j = max{TRC k,l-1,l + TI k,l,i,j , TC i,j-1};

[0027] where i, p, k are the i-th, p-th, and k-th workpieces respectively, and the range of the workpiece is i = 1, 2,..., n; p = 1, 2,..., n; k = 1, 2,..., n; j, q, l are the processes corresponding to the i-th, p-th, and k-th workpieces, and the range of the process is j = 1, 2,..., o i ; q = 1, 2,..., o i ; l = 1, 2,..., o i ; TS i,j represents the processing start time of process O ij , TWC i,0 represents the end time of the transportation task before O i1 , TRC i,j-1,j represents the end time of the transportation task before O ij-1 starts after the O ij process of the workpiece is completed, and TCp,q Represents the previous workpiece O on the machine pq End time of the process, m i,j Represents process O ij Selected processing machine; TRS i,j-1,j Represents O of workpiece i ij-1 After the process is completed, O ij Start time of the transportation task before the process starts, TRC k,l-1,l Represents the previous transportation task R of the rail-guided vehicle RGV kl Transportation end time, TI k,l,i,j Represents the completion of R kl After the transportation task, go to the next transportation task R ij Idle transportation time, TC i,j-1 Represents O ij The previous process of O ij-1 End time of processing on the machine.

[0028] Preferably, in the flexible intelligent manufacturing unit warehouse, the specific expression for the transportation time between each storage location and each machine is:

[0029] t (i,j),k = s (i,j),k / v;

[0030] In the formula, i is the layer number of the workpiece in the warehouse, and the range of change is i = 1, 2,..., H; j is the column number of the workpiece in the warehouse, and its range of change is i = 1, 2,..., L; k is the machine number, and its range of change is k = 1, 2,..., m; t (i,j),k Represents the transportation time between the (i, j) position in the warehouse and machine k; s (i,j),k Represents the transportation distance between the (i, j) position in the warehouse and machine k; v represents the average transportation speed of the RGV;

[0031] The specific expression for the transportation time between each machine and other machines is:

[0032] t k1,k2 = s k1,k2 / v;

[0033] In the formula, k1, k2 are machine numbers, and their range of change is k1, k2 = 1, 2,..., m; t k1,k2 Represents the transportation time between machine k1 and machine k2; s k1,k2 Represents the transportation distance between machine k1 and machine k2; v represents the average transportation speed of the RGV.

[0034] Preferably, the genetic algorithm includes a crossover operation and a mutation operation, specifically:

[0035] The crossover operation includes three operation operators. Among them, operation operator 1 performs crossover operation on the process coding layer; operation operator 2 operates on the machine selection coding layer; operation operator 3 operates on the position coding layer of the workpiece in the warehouse. The scheduling scheme is processed through the crossover operation;

[0036] The mutation operation includes three operation operators. Operation operator 1 performs mutation operation on the process coding layer; operation operator 2 operates on the machine selection coding layer; operation operator 3 operates on the position coding layer of the workpiece in the warehouse. A workpiece in the matrix is randomly selected, and a vacant position is randomly selected to place the workpiece, and the original position of the workpiece is set as vacant.

[0037] Preferably, performing iterative variable neighborhood search on the high-quality solution includes:

[0038] When forming a key block with the processing operation time, the head and tail adjacent load times in the key block are exchanged to generate a new neighborhood solution 1, and the corresponding makespan1 of the completion time is calculated;

[0039] The head and tail adjacent load times of the key block are corresponding to the adjacent processing operations in the coding. The positions of the two processing operations are exchanged in the coding, and the neighborhood solution 2 is decoded, and the corresponding makespan2 of the completion time is calculated;

[0040] The completion time of the neighborhood solution 2 is compared with the completion time of the optimal solution before the position exchange operation. If the completion time of the neighborhood solution 2 is less than the completion time of the optimal solution before the exchange operation, the coding corresponding to the neighborhood solution 2 is used as the starting point for the next iteration; wherein the completion time of the optimal solution is the minimized makespan;

[0041] The completion time of the neighborhood solution 1 is compared with the completion time of the optimal solution before the position exchange operation and the completion time of the neighborhood solution 2. If the completion time of the neighborhood solution 1 is the smallest, the neighborhood solution 1 and the completion time of the neighborhood solution 1 are used as a solution and saved for subsequent comparison.

[0042] The present invention provides a collaborative optimization scheduling system for the warehouse and processing of a flexible manufacturing cell, including:

[0043] An acquisition module for acquiring flexible operation-related data and the layout of machines and warehouse storage locations in the flexible intelligent manufacturing cell;

[0044] A model optimization module for considering the flexible processing scheduling in the flexible intelligent manufacturing cell and the scheduling of the workpiece in the warehouse storage location, and constructing a scheduling model for the collaborative optimization of the warehouse and processing of the flexible intelligent manufacturing cell with the minimization of the makespan as the optimization goal;

[0045] A scheduling scheme acquisition module is used to input flexible job-related data and the layout of machines and storage locations into a scheduling model, obtain a feasible solution including the processing sequence of processes, the processing machines selected for the processes, and the locations of workpieces in the warehouse, decode the feasible solution to obtain a scheduling scheme for the corresponding flexible job shop and the locations of workpieces in the warehouse, process this scheduling scheme through a genetic algorithm, select the solution with the minimum makespan as the high-quality solution, perform iterative variable neighborhood search on the high-quality solution, and when the number of iterations is satisfied, output the optimal individual in the population, and obtain the scheduling scheme for the processing and storage of the flexible intelligent manufacturing unit according to the optimal individual in the population.

[0046] The present invention provides a computer device, including a memory and a processor. A program is stored in the memory. When the program is executed by the processor, the processor executes the steps of the above-mentioned collaborative optimization scheduling method for the storage and processing of a flexible manufacturing unit.

[0047] Compared with the prior art, the present invention has the following remarkable advantages:

[0048] According to the characteristics of the flexible intelligent manufacturing unit, the present invention establishes an integrated optimization scheduling model for storage and processing, performs collaborative scheduling on the storage and processing processes, comprehensively considers the location layout of workpieces in the warehouse, the coordination of process scheduling and transportation tasks, obtains a reasonable storage scheme and an efficient transportation path and transportation time while optimizing the production scheduling scheme, optimizes the storage resources, combines variable neighborhood search and genetic algorithm to solve the scheduling problem, effectively combines the encoding and decoding of the locations of workpieces in the warehouse with the neighborhood structure to ensure jumping out of the local optimum, further expands the optimization space of the scheduling solution, can explore more high-quality solutions, significantly improves the scheduling optimization effect, and realizes the overall optimum of the space-time efficiency of material flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is the overall scheme flowchart provided by the present invention;

[0050] Figure 2 is the schematic diagram of the three-layer coding scheme of the present invention;

[0051] Figure 3 is the Gantt chart of the preferred embodiment of the present invention; wherein Figure 3 the (a) of which is the Gantt chart for the first 250, Figure 3 the (b) of which is the Gantt chart after 250;

[0052] Figure 4 is the schematic diagram of the crossover operation operator 3 involved in the present invention.

[0053] Figure 5 is the flowchart of the improved process exchange method involved in the present invention;

[0054] Figure 6 This is a flowchart of a method for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell of the present invention. Specific embodiments

[0055] Next, in conjunction with the accompanying drawings in the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0056] The method first performs scheduling modeling on a flexible job shop considering warehousing, and uses a hybrid optimization algorithm combining genetic algorithm and variable neighborhood search to optimize and solve the obtained scheduling model, so that the production scheduling and warehousing scheduling in the flexible intelligent manufacturing cell including warehousing are optimized, the transportation time can be reasonably optimized, the production efficiency can be improved, the production cycle can be shortened, and greater economic benefits can be brought.

[0057] Specifically, the method first needs to construct a collaborative optimization scheduling model for warehousing and processing in a flexible intelligent manufacturing cell. Secondly, a hybrid optimization algorithm combining genetic algorithm and variable neighborhood search is proposed to optimize and solve the obtained scheduling model. It is necessary to simultaneously consider the flexible processing scheduling and warehousing scheduling in the workshop, and the optimization goal is to minimize the makespan. The makespan is the maximum value of the completion times of the last operations of all workpieces.

[0058] As Figure 1 and Figure 6 shown, a method for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell in this embodiment includes the following steps:

[0059] Step S1: Collect data related to flexible operations and the layout of machines and warehousing locations in the flexible intelligent manufacturing cell.

[0060] The flexible intelligent manufacturing cell consists of an automated storage and retrieval system (AS / RS), a rail-guided vehicle (RGV), an RGV track, and a numerically controlled machine tool. The RGV track is a straight track, and the warehousing is a side-of-line warehouse parallel to the RGV track, with a row of H-layer L-column shelves. An RGV moves on the RGV track, and there is a robotic arm on the RGV to grab workpieces at different positions in the warehousing. The RGV moves to the position corresponding to the machine and places the workpiece in the buffer area in front of the machine. The specific expression for the transportation time between each storage location in the warehousing and each machine is:

[0061] t (i,j),k = S (i,j),k / v;

[0062] where \(i\) is the number of layers of workpieces in the warehouse, and its range of variation is \(i = 1, 2, \ldots, H\); \(j\) is the number of columns of workpieces in the warehouse, and its range of variation is \(i = 1, 2, \ldots, L\); \(k\) is the machine number, and its range of variation is \(k = 1, 2, \ldots, m\); \(t\) (i,j),k represents the transportation time between the position \((i, j)\) in the warehouse and machine \(k\); \(S\) (i,j),k represents the transportation distance between the position \((i, j)\) in the warehouse and machine \(k\); \(v\) represents the average transportation speed of the RGV.

[0063] The specific expression of the transportation time between each machine and other machines is as follows:

[0064] t k1,k2 = s k1,k2 / v;

[0065] where \(k1\), \(k2\) are machine numbers, and their range of variation is \(k1, k2 = 1, 2, \ldots, m\); \(t\) k1,k2 represents the transportation time between machine \(k1\) and machine \(k2\); \(s\) k1,k2 represents the transportation distance between machine \(k1\) and machine \(k2\); \(v\) represents the average transportation speed of the RGV.

[0066] Organize the transportation times between the warehouse and machines, and between machines into a transportation time table, which is used as one of the input data for the optimization algorithm.

[0067] Step S2: Considering the flexible machining scheduling in the flexible intelligent manufacturing cell and the scheduling of the storage locations of workpieces in the warehouse, with the minimization of the makespan as the optimization goal, construct a scheduling model for the collaborative optimization of the warehouse and processing in the flexible intelligent manufacturing cell.

[0068] The scheduling model is as follows: There are \(m\) processing machines, \(n\) workpieces to be processed, a shelf with \(H\) layers and \(L\) columns in the warehouse, and a transportation device in the workshop for transportation tasks from the warehouse to machines and between machines. Each workpiece contains \(o\) i processes, and each process selects a machine for processing. At time 0, each workpiece is located at a certain position in the warehouse, and the transportation device RGV is located at a given initial position. There is a transportation time before the first process of the workpiece starts, and there is a transportation time between adjacent processes of the same workpiece.

[0069] The specific expressions of the processing start time and transportation start time of each process of each workpiece in the scheduling model are as follows:

[0070] TS i,1 = max{TWC i,0 , TC p,q};

[0071]

[0072] TRS i,j-1,j = max{TRC k,l-1,l + TI k,l,i,j , TC i,j-1};

[0073] Wherein, i, p, and k are the i-th, p-th, and k-th workpieces respectively, and the range of workpieces is i = 1, 2,..., n; p = 1, 2,..., n; k = 1, 2,..., n; j, q, and l are the processes corresponding to the i-th, p-th, and k-th workpieces, and the range of processes is j = 1, 2,..., o i ; q = 1, 2,..., o i ; l = 1, 2,..., o i ; TS i,j represents the start time of processing of process O ij TWC i,0 represents the end time of the transportation task before the start of O i1 TRC i,j-1,j represents the end time of the transportation task before the start of O ij-1 after the completion of the O process of the workpiece, and before the start of the O ij process, TC p,q represents the end time of the processing of the previous workpiece Q pq on the machine, m i,j represents the processing machine selected for process O ij ; TRS i,j-1,j represents the start time of the transportation task before the start of O ij-1 after the completion of the O process of workpiece i, and before the start of the O ij process, TRC k,l-1,l represents the end time of the previous transportation task R kl of the Rail Guided Vehicle RGV, TI k,l,i,j represents the no-load transportation time to the next transportation task R kl after completing the R ij transportation task, TC i,j-1 represents the end time of the previous process O ij of O ij-1 on the machine. When the current process of the workpiece and the previous process of the workpiece are processed on the same machine, the start time of processing is equal to the end time of the processing of the previous workpiece on the machine; otherwise, the start time of processing is equal to the maximum value of the end time of the transportation task of this process and the end time of the processing of the previous workpiece on the machine. The start time of the transportation task is equal to the end time of the previous transportation task of the RGV plus the maximum value of the no-load movement time of this task and the end time of the previous process of this task on the machine.

[0074] The specific expression of the makespan in the scheduling model is:

[0075]

[0076] Among them, C max is the makespan, is the completion time of operation O ij .

[0077] The optimization objective is to determine the processing sequence of all processing operations of n workpieces, the selection of machines, and the positions of n workpieces in the warehouse, so as to minimize the makespan C max .

[0078] Step S3: Input the flexible job-related data, as well as the machine and warehouse location layout, into the scheduling model to obtain a feasible solution including the processing sequence of operations, the processing machines selected for the operations, and the positions of the workpieces in the warehouse. Decode the feasible solution to obtain the corresponding scheduling plan for the flexible job shop and the positions of the workpieces in the warehouse, and process this scheduling plan through a genetic algorithm. Select the solution with the minimum makespan as the high-quality solution, perform iterative variable neighborhood search on the high-quality solution. When the number of iterations is satisfied, output the optimal individual of the population, and obtain the scheduling plan for the processing and storage of the flexible intelligent manufacturing unit according to the optimal individual of the population.

[0079] As Figure 2 shown, a three-layer coding form is used to represent a feasible solution of the scheduling model. Decoding the feasible solution gives the corresponding scheduling plan for the flexible job shop and the positions of the workpieces in the warehouse. Specifically:

[0080] Coding: Use the processing sequence of all obtained operations as the first layer, that is, the operation coding layer; use the processing machines selected for each operation as the second layer, that is, the machine selection coding layer; use the positions of all workpieces in the warehouse as the third layer, that is, the position coding layer of the workpieces in the warehouse. Represent the feasible solution in the form of the operation coding layer, the machine selection coding layer, and the position coding layer of the workpieces in the warehouse. The specific expression of the position coding Z of the workpieces in the warehouse is:

[0081] Z = [Z(l, h)] H×L ;

[0082] where represents the state of this storage unit.

[0083] Decoding: First, obtain the processing sequence of all processes according to the process coding layer, and sort them in sequence. Then, obtain the processing machine and the corresponding processing time of this process according to the machine selection coding layer. When processing the first process of the workpiece, obtain the position of the workpiece in the warehouse according to the position coding layer of the workpiece in the warehouse, so as to obtain the transportation time between the warehouse and the machine. When processing the subsequent processes of the workpiece, calculate the transportation time between different machines. Judge the no-load start time of the RGV, calculate the load transportation start time and the load transportation end time. Finally, judge the processing start time of this process, calculate the processing end time of this process. After processing in sequence according to the process coding layer, obtain the processing end time of all processes, and calculate the makespan C max 。

[0084] Initialization of the population. The population size is PopSize, and the random generation method is used to initialize each individual of the population. Execute the encoding and decoding method for each individual in the population. As a schematic preference, in this embodiment, PopSize is taken as 100.

[0085] Setting of hybrid algorithm parameters. Set the maximum number of iterations Gen of the genetic algorithm max 、crossover probability, mutation probability, the maximum number of iterations t of the variable neighborhood search algorithm max 、the number i of neighborhood structures of the variable neighborhood search algorithm max 。As a schematic preference, in this embodiment, Gen max is taken as 200, the crossover probability is taken as 0.9, and the mutation probability is taken as 0.15.

[0086] Selection operation. Adopt a two-stage selection strategy to screen the population. First, perform elite selection. Sort according to the makespan obtained by decoding, and directly retain some individuals with a short makespan to ensure the inheritance of high-quality solutions. Then, use tournament selection for the remaining individuals. Randomly select two individuals for comparison each time, and select the better individual to enter the next generation population. This strategy enhances the diversity of the population while ensuring the retention of high-quality individuals, thereby improving the global search ability of the algorithm.

[0087] The crossover operation includes three operation operators, and the specific operations of the three operators are executed for each individual undergoing crossover in this embodiment.

[0088] Operation Operator 1: Perform crossover operation on the process coding layer. Randomly select one of the POX and JBX operators for crossover operation on the two individuals to be crossed. Assume that the process codes of the current two individuals are [1, 3, 1, 2, 2, 3] and [3, 2, 1, 2, 3, 1] respectively, and the workpieces are randomly assigned into two groups {2} and {1, 3}. If the POX operation is performed, the two individuals obtained are [3, 1, 3, 2, 2, 1] and [1, 2, 3, 2, 1, 3]; if the JBX operation is performed, the two individuals obtained are [3, 1, 3, 2, 2, 1] and [3, 2, 1, 2, 3, 1].

[0089] Operation Operator 2: Operate on the machine selection coding layer. Randomly generate a crossover start point and end point, and exchange the machine selection codes of the two individuals within this interval.

[0090] Operation Operator 3: Operate on the position coding layer of the workpieces in the warehouse. First, select a sub-rectangular area, and randomly select the start point (startX, startY) and end point (endX, endY); then exchange the values of the two parent matrices in this area. Since after the exchange, it may cause a workpiece to appear twice or be missing in the new matrix, reallocation is required; perform workpiece position correction. Traverse the entire matrices of the two newly generated matrices, find the duplicate workpieces, remove the duplicate workpieces outside the sub-rectangular area, and keep the workpieces within the area. Then traverse the matrices to count which workpieces are missing, and randomly select empty positions for filling. Assume that the position codes of the workpieces of the current two individuals in the warehouse are [[1, 0, 3, 10, 0, 7], [4, 5, 8, 0, 0, 0], [6, 2, 0, 0, 9, 0]] and [[4, 0, 0, 9, 0, 0], [1, 2, 3, 5, 0, 0], [8, 6, 10, 0, 7, 0]] respectively, and the exchange start point and end point are (0, 0) and (1, 3) respectively, then the codes of the obtained individuals are [[4, 0, 8, 9, 10, 0], [1, 2, 3, 5, 0, 0], [6, 0, 0, 0, 0, 0]] and [[1, 0, 3, 10, 0, 0], [4, 5, 8, 0, 0, 0], [2, 6, 0, 9, 7, 0]], as Figure 4 shown.

[0091] The mutation operation includes three operation operators, and the specific operations of the three operators are performed on each individual that undergoes crossover in this embodiment.

[0092] Operation Operator 1: Perform mutation operation on the process coding layer; Operation Operator 2: Perform mutation operation on the machine selection coding layer; Operation Operator 3: Operate on the position coding layer of the workpieces in the warehouse. First, randomly select a workpiece in the matrix; then randomly select an empty position to place this workpiece; finally, set the original position of the workpiece as empty.

[0093] Perform variable neighborhood search (VNS) on the high-quality solution. The specific implementation steps are as follows:

[0094] (1) Set the neighborhood i = 1.

[0095] (2) Set the iteration number t = 0.

[0096] (3) Conduct local search in the neighborhood structure i, t = t + 1.

[0097] (4) If a better solution is found through the search, execute step (5); otherwise, execute step (7).

[0098] (5) Replace the current optimal solution with the better solution obtained through the search.

[0099] (6) If the iteration number t is greater than t max , then execute step (7); otherwise, return to step (3).

[0100] (7) If the neighborhood i is equal to i max , then output the current optimal solution; otherwise, execute i = i + 1 and then return to step (2).

[0101] It includes three neighborhood structures. Neighborhood structure 1: improved N5 neighborhood structure. First, identify the critical path in the current scheduling plan. Next, adjust the operations on the critical path. When the critical block consists of processing operation times, swap the positions of the two adjacent processing operations at the head and tail of the critical block; when the critical block consists of transportation times, the neighborhood solution obtained by simply swapping adjacent loads cannot be directly converted into operation encoding. As shown in Figure 5 , adopt an improved operation swapping method, which includes the following sub-steps:

[0102] (1) Swap the load times of the two adjacent loads at the head and tail of the critical block to generate a new neighborhood solution (denoted as neighborhood solution 1), and calculate its corresponding makespan makespan1.

[0103] (2) Correlate the load times of the two adjacent loads at the head and tail of the critical block to the adjacent processing operations in the encoding, swap the positions of these two processing operations in the encoding, and decode to generate a new neighborhood solution (denoted as neighborhood solution 2), and calculate its corresponding makespan makespan2.

[0104] (3) Compare the makespan makespan2 of neighborhood solution 2 with the makespan of the optimal solution before the swapping operation. If makespan2 is less than the original optimal makespan, it indicates that neighborhood solution 2 has potential in the improvement direction. Therefore, use the encoding corresponding to neighborhood solution 2 as the starting point for the next iteration and continue to optimize.

[0105] (4) Compare the makespan1 of the neighborhood solution 1 with the makespan2 of the optimal solution before the swapping operation. If makespan1 is the smallest, although it cannot be directly converted into process coding for iteration, save the neighborhood solution 1 and its corresponding makespan1 as a better solution for subsequent comparison.

[0106] Neighborhood structure 2: A neighborhood structure based on different machines. For the processes on the critical path, change the selected machine in the coding and randomly select from the alternative machines.

[0107] Neighborhood structure 3: A neighborhood structure based on different storage locations. Randomly select a workpiece in the warehouse and change its location.

[0108] If the algorithm satisfies that the current iteration number is equal to Gen max , output the optimal individual of the population; otherwise, jump to the step and iterate repeatedly until the termination condition is met.

[0109] The present invention proposes a collaborative optimization scheduling system for the storage and processing of a flexible manufacturing cell, including: a collection module, a model optimization module, and a scheduling scheme acquisition module.

[0110] Among them, the collection module is used to collect flexible operation-related data and the layout of machines and storage locations in the flexible intelligent manufacturing cell; the model optimization module is used to consider the flexible processing scheduling in the flexible intelligent manufacturing cell and the scheduling of workpiece storage locations, and construct a scheduling model for the collaborative optimization of the storage and processing of the flexible intelligent manufacturing cell with the goal of minimizing the makespan; the scheduling scheme acquisition module is used to input the flexible operation-related data and the layout of machines and storage locations into the scheduling model, obtain a feasible solution including the processing sequence of processes, the processing machines selected by the processes, and the locations of workpieces in the warehouse, decode the feasible solution to obtain a scheduling scheme for the corresponding flexible job shop and the locations of workpieces in the warehouse, and process the scheduling scheme through a genetic algorithm, select the solution with the minimum makespan as the high-quality solution, perform iterative variable neighborhood search on the high-quality solution, and when the iteration number is satisfied, output the optimal individual of the population, and obtain the scheduling scheme for the processing and storage of the flexible intelligent manufacturing cell according to the optimal individual of the population.

[0111] The present invention also provides a computer device, including a memory and a processor. When a program stored in the memory is executed by the processor, the processor executes the steps of a collaborative optimization scheduling method for the storage and processing of a flexible manufacturing cell.

[0112] According to the disclosed embodiments, a computer device may communicate with one or more external devices (e.g., a keyboard, a pointing device, Bluetooth communication, etc.), or with any device (e.g., a router, a demodulator, etc.) that enables the computing device to communicate with one or more other computing devices.

[0113] According to the above steps, in this embodiment, a case study with 6 workpieces and 6 machines is selected for experiments. The optimal scheduling plan obtained by this algorithm in 10 independent experiments is the plan with a makespan of 378, as Figure 3 shown.

[0114] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. For those skilled in the technical field to which the present invention pertains, without departing from the concept of the present invention, several simple deductions or substitutions can be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. A method for collaborative optimization scheduling of warehousing and processing in a flexible manufacturing cell, characterized in that, Including: Collecting data related to flexible operations and the layout of machines and storage locations in the flexible intelligent manufacturing unit; Considering the flexible machining scheduling in the flexible intelligent manufacturing unit and the scheduling of workpiece storage locations, and constructing a scheduling model for the collaborative optimization of storage and machining in the flexible intelligent manufacturing unit with the objective of minimizing the makespan; Inputting the data related to flexible operations and the layout of machines and storage locations into the scheduling model, obtaining a feasible solution including the processing sequence of operations, the processing machines selected for the operations, and the positions of the workpieces in the warehouse, decoding the feasible solution to obtain a scheduling plan for the corresponding flexible job shop and the positions of the workpieces in the warehouse, and processing the scheduling plan through a genetic algorithm, selecting the solution with the minimum makespan as the high-quality solution, performing iterative variable neighborhood search on the high-quality solution, and when the number of iterations is satisfied, outputting the optimal individual of the population, and obtaining the scheduling plan for the machining and storage of the flexible intelligent manufacturing unit according to the optimal individual of the population.

2. The warehousing and processing collaborative optimization scheduling method for a flexible manufacturing cell according to claim 1, characterized in that, The obtaining of the feasible solution including the processing sequence of operations, the processing machines selected for the operations, and the positions of the workpieces in the warehouse is specifically as follows: Taking the processing sequence of all obtained operations as the first layer, that is, the operation coding layer; taking the processing machines selected for each operation as the second layer, that is, the machine selection coding layer; taking the positions of all workpieces in the warehouse as the third layer, that is, the position coding layer of the workpieces in the warehouse; representing the feasible solution in the form of the operation coding layer, the machine selection coding layer, and the position coding layer of the workpieces in the warehouse, where the specific expression of the position coding Z of the workpieces in the warehouse is: Z = [Z(h, l)] H×L ; wherein represents the state of the memory cell.

3. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 2, characterized in that The decoding of the feasible solution to obtain the corresponding scheduling plan for the flexible job shop and the positions of the workpieces in the warehouse includes: Obtaining the processing sequence of all operations according to the operation coding layer, sorting them in sequence according to the processing sequence, and obtaining the processing machine and the corresponding processing time of the operation according to the machine selection coding layer; When processing the first operation of the workpiece, obtaining the position of the workpiece in the warehouse according to the position coding layer of the workpiece in the warehouse, obtaining the transportation time between the warehouse and the machine, and when processing the subsequent operations of the workpiece, calculating the transportation time between different machines; Determine the no-load start time of the RGV, obtain the load transportation start time and load transportation end time, and determine the processing start time of this process, calculate the processing end time of this process, and after processing in the order of the process coding layer, obtain the processing end times of all processes with the load transportation start time, load transportation end time and the processing start times of all processes, and calculate the makespan C with the load transportation start time and the processing end times of all processes max ; The load is the relationship between workpieces in storage and machines and the relationship between workpieces in different machines.

4. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 3, wherein, The specific expression of the makespan is: Among them, C max is the makespan; is the completion time of operation O ij , that is, the sum of the processing time and transportation time of each operation.

5. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 4, characterized in that, Obtaining the processing time and transportation time of each operation of each workpiece in the scheduling model, and the specific expression is: TS i,1 = max{TWC i,0 , TC p,q}; TRS i,j-1,j = max{TRC k,l-1,l + TI k,l,i,j , TC i,j-1}; Among them, i, p, and k are the i-th, p-th, and k-th workpieces respectively, and the range of workpieces is i = 1, 2, …, n; p = 1, 2, …, n; k = 1, 2, …, n; j, q, and l are the processes corresponding to the i-th, p-th, and k-th workpieces respectively, and the range of processes is j = 1, 2, …, o i ; q = 1, 2, …, o i ; l = 1, 2, …, o i ; TS i,j represents the start time of processing for the O ij process, TWC i,0 represents the end time of the transportation task before the start of O i1 start, TRC i,j-1,j represents the O of the workpiece ij-1 After the process is completed, O ij the end time of the transportation task before the start of the O process, TC p,q represents the previous workpiece O on the machine pq the end time of the process machining, m i,j represents O ij the processing machine selected for the process; TRS i,j-1,j represents the O of workpiece i ij-1 After the process is completed, O ij the start time of the transportation task before the start of the O process, TRC k,l-1,l represents the end time of the previous transportation task R of the Rail Guided Vehicle RGV kl transportation, TI k,l,i,j represents the completion of R kl After the transportation task, go to the next transportation task R ij the no-load transportation time, TC i,j-1 represents O ij the previous process O of ij-1 the end time of the machining on the machine.

6. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 3, wherein, In the storage of the flexible intelligent manufacturing unit, the specific expression of the transportation time between each storage location and each machine is: t (i,j),k = s (i,j),k / v; In the formula, i is the layer number of the workpiece in the warehouse, and the range of change is i = 1, 2,..., H; j is the column number of the workpiece in the warehouse, and the range of change is i = 1, 2,..., L; k is the number of the machine, and the range of change is k = 1, 2,..., m; t (i,j),k represents the transportation time from the (i, j) position in the warehouse to machine k; s (i,j),k represents the transportation distance between the (i, j) position in the warehouse and machine k; v represents the average transportation speed of the RGV; The specific expression of the transportation time between each machine and other machines is: t k1,k2 = s k1,k2 / v; In the formula, k1, k2 are the numbers of the machines, and the range of change is k1, k2 = 1, 2,..., m; t k1,k2 represents the transportation time between machine k1 and machine k2; s k1,k2 represents the transportation distance between machine k1 and machine k2; v represents the average transportation speed of the RGV.

7. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 2, characterized in that, The genetic algorithm includes crossover operation and mutation operation, and specifically: The crossover operation includes three operation operators. Among them, operation operator 1 performs crossover operation on the process coding layer; operation operator 2 operates on the machine selection coding layer; operation operator 3 operates on the position coding layer of the workpiece in the warehouse, and processes the scheduling scheme through the crossover operation; The mutation operation includes three operation operators. Operation operator 1 performs mutation operation on the process coding layer; operation operator 2 operates on the machine selection coding layer; operation operator 3 operates on the position coding layer of the workpiece in the warehouse. Randomly select a workpiece in the matrix and randomly select a vacant position to place the workpiece, and set the original position of the workpiece as vacant.

8. The warehousing and processing collaborative optimization scheduling method of a flexible manufacturing cell according to claim 1, characterized in that, Performing iterative variable neighborhood search on the high-quality solution includes: When forming a key block with the processing operation time, exchange the head and tail adjacent load times in the key block to generate a new neighborhood solution 1, and calculate the corresponding makespan1; Correspond the head and tail adjacent load times of the key block to the adjacent processing operations in the coding, exchange the positions of the two processing operations in the coding, and decode to generate neighborhood solution 2, and calculate the corresponding makespan2; Compare the makespan of neighborhood solution 2 with the makespan of the optimal solution before the position exchange operation. If the makespan of neighborhood solution 2 is less than the makespan of the optimal solution before the exchange operation, then use the coding corresponding to neighborhood solution 2 as the starting point for the next iteration; where the optimal solution makespan is the minimized maximum completion time; Compare the makespan of neighborhood solution 1 with the makespan of the optimal solution before the position exchange operation and the makespan of neighborhood solution 2. If the makespan of neighborhood solution 1 is the smallest, then use neighborhood solution 1 and the makespan of neighborhood solution 1 as a solution and save it for subsequent comparison.

9. A warehousing and processing collaborative optimization scheduling system for a flexible manufacturing cell, characterized in that, Including: A collection module for collecting flexible operation-related data and the layout of machines and warehouse storage locations in the flexible intelligent manufacturing unit; A model optimization module for considering the flexible processing scheduling in the flexible intelligent manufacturing unit and the scheduling of the workpiece in the warehouse storage location, with the goal of minimizing the maximum completion time, and constructing a scheduling model for the collaborative optimization of the warehouse and processing in the flexible intelligent manufacturing unit; A scheduling scheme acquisition module for inputting flexible operation-related data and the layout of machines and warehouse storage locations into the scheduling model, obtaining a feasible solution including the processing sequence of the operations, the processing machines selected for the operations, and the positions of the workpieces in the warehouse, decoding the feasible solution to obtain the scheduling scheme of the corresponding flexible job shop and the positions of the workpieces in the warehouse, and processing the scheduling scheme through a genetic algorithm, selecting the solution with the minimized maximum completion time as the high-quality solution, performing iterative variable neighborhood search on the high-quality solution, and when the iteration times are met, outputting the optimal individual of the population, and obtaining the processing and warehouse scheduling scheme of the flexible intelligent manufacturing unit according to the optimal individual of the population.

10. A computer device, characterized in that, Including a memory and a processor. When the program stored in the memory is executed by the processor, the processor executes the steps of a method for collaborative optimization scheduling of warehouse and processing in a flexible manufacturing unit as described in any one of claims 1 to 8.

Images