Customized production and boxing collaborative scheduling method and system, storage medium and product

Abstract

The application provides a customized production and boxing collaborative scheduling method and system, a storage medium and a product. The method comprises the following steps: S1, obtaining an information matrix and a delivery period of workpieces of each delivery address; S2, using a problem-specific knowledge method to generate strategy parameters and multiple sets of collaborative scheduling schemes according to the information matrix and the delivery period of the workpieces of each delivery address; S3, performing non-dominated sorting on each set of collaborative scheduling schemes based on multiple target values obtained by inputting each candidate scheduling scheme into a production and boxing collaborative scheduling model, and establishing a hierarchical storage Archive population; and S4, updating the strategy parameters, selecting a collaborative scheduling representative scheme from the hierarchical storage Archive population through a roulette method, and adjusting other candidate collaborative scheduling schemes according to a fitness value corresponding to the collaborative scheduling representative scheme, so that high production efficiency and high boxing efficiency of customized orders can be considered.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent logistics, and in particular to a customized production and boxing collaborative scheduling method and system, a storage medium and a product. BACKGROUND

[0002] In today's competitive market environment, the customized production mode has become the key strategy for many manufacturing enterprises to meet customers' individualized needs and improve market competitiveness. In the clothing production industry, for example, making clothes according to customers' needs has become the development direction of the industry. With the increasing demand for response speed and delivery flexibility of individualized orders, the traditional isolated optimization operation mode has been difficult to meet the efficiency needs of the whole supply chain, and it is urgent to realize efficient collaborative production throughout the chain.

[0003] In the customized production process, there are multiple production, quality inspection, warehousing and material boxing links, and the main core links affecting the whole process efficiency are production and logistics boxing. The planning and scheduling of the production link determines the procurement, input of raw materials and output of products, and the logistics boxing affects the collaborative scheduling of multiple different products. The manufacturing system needs to integrate the collaborative decision of production scheduling and material boxing to realize the whole link optimization from order processing to product delivery.

[0004] Currently, production and product boxing are independently scheduled, and the degree of production collaboration is low. In the field of production scheduling, genetic algorithms, mixed integer programming and other methods are used to optimize equipment load and order scheduling according to production process requirements. For example, in the paper "Research on AGV and Production Integrated Scheduling Considering Order Batching" (Management and Management, 2025), Ruisyu and Han Wenmin considered the order batching and workpiece transportation time factors and established an AGV and production integrated scheduling model. However, this model only optimizes the production scheduling stage. Meanwhile, the material boxing in other methods only considers the optimization of boxing efficiency, box volume rate and other targets.

[0005] The optimization strategies of production and product boxing are independent of each other. Although good optimization results have been achieved in each link, the lack of system collaboration caused by isolated optimization has seriously restricted the overall efficiency. For example, in the phased delivery scenario, the existing mode is difficult to coordinate the timing relationship between production rhythm and boxing batch, often leading to delayed shipment of completed orders due to waiting for boxing, or forced early production to meet transportation batch, resulting in increased inventory costs; in order to meet the needs of some high-quality customers for fast delivery, production planning is often disrupted, causing production chaos. This conflict between links not only causes resource waste, but also directly affects the order fulfillment ability of enterprises and customer satisfaction.

[0006] Therefore, the existing optimization method cannot meet the market demand of modern manufacturing enterprises in terms of high production efficiency and high packing efficiency of customized orders, and there is an urgent need for a collaborative scheduling method that can comprehensively consider various factors to improve the current situation and maintain customer satisfaction. SUMMARY

[0007] The application provides a customized production and packing collaborative scheduling method, system, storage medium and product, which can consider high production efficiency and high packing efficiency of customized orders.

[0008] The first aspect of the application provides a customized production and packing collaborative scheduling method, which comprises:

[0009] Step S1: obtaining an information matrix and a delivery period of each workpiece of each delivery address, the information matrix comprising an order demand matrix, a box capacity matrix, a processing time matrix and a workpiece occupied capacity matrix;

[0010] Step S2: generating strategy parameters and multiple groups of collaborative scheduling schemes according to the information matrix and the delivery period of each workpiece of each delivery address by using a problem-specific knowledge method, wherein each group of collaborative scheduling schemes comprises multiple alternative scheduling schemes of the same number of boxes;

[0011] Step S3: performing non-dominated sorting on each group of collaborative scheduling schemes based on multiple target values obtained by inputting each alternative scheduling scheme into a production and packing collaborative scheduling model, and establishing a hierarchical storage Archive population;

[0012] Step S4: updating the strategy parameters, selecting a collaborative scheduling representative scheme from the hierarchical storage Archive population by using a roulette method, and adjusting other alternative collaborative scheduling schemes according to the fitness value corresponding to the collaborative scheduling representative scheme;

[0013] Step S5: adjusting each alternative collaborative scheduling scheme based on a neighborhood search method;

[0014] Step S6: determining whether the adjusted alternative collaborative scheduling scheme meets a preset constraint condition, if not, executing step S5, and if yes, executing step S7;

[0015] Step S7: performing non-dominated sorting on each group of collaborative scheduling schemes based on the fitness value of each alternative collaborative scheduling scheme, to update the hierarchical storage Archive population;

[0016] Step S8: determining whether the number of the Archive population reaches a preset upper limit, if yes, removing the alternative collaborative scheduling scheme with a congestion degree greater than a preset congestion degree, and if no, executing step S9;

[0017] Step S9: judging whether the maximum iteration number is reached, otherwise re-executing step S4, yes outputting the collaborative scheduling scheme group, determining the target collaborative scheduling scheme from the multiple collaborative scheduling scheme groups based on the multiple target values of the current iteration.

[0018] In some embodiments of the first aspect, the production and boxing collaborative scheduling model is constructed to maximize average boxing volume utilization rate, minimize maximum completion time of multiple order information, and minimize early / late cost, and is constructed in combination with multi-objective optimization theory.

[0019] In some embodiments of the first aspect, the multiple target values include average boxing volume utilization rate, maximum completion time, and early / late cost, and step S3 includes:

[0020] determining total address quantity of each consignee address in the alternative scheduling scheme, total capacity of the box corresponding to each consignee address, and actual occupied capacity of the workpiece in the box;

[0021] determining boxing volume utilization rate of each consignee address according to the total capacity of the box and the actual occupied capacity of the workpiece in the box;

[0022] obtaining average boxing volume utilization rate according to the boxing volume utilization rate and the total address quantity of each consignee address; and

[0023] determining completion time of the last workpiece in each consignee address in the alternative scheduling scheme;

[0024] accumulating the last completion time to obtain the maximum completion time; and

[0025] obtaining early completion time and late time of each consignee address according to the delivery period of the workpiece required by each consignee address in the alternative scheduling scheme and the completion time of the last workpiece in each consignee address;

[0026] obtaining early / late cost according to the early completion time and the late time, and preset early completion unit cost and late unit cost.

[0027] In some embodiments of the first aspect, step S2 includes:

[0028] according to the information matrix and the delivery period of the workpiece of each consignee address, combining the greedy and random strategies, and respectively dividing the type corresponding to each type of box and the corresponding condition of the quantity of different workpieces of each consignee address corresponding to each type of box into corresponding box type codes and boxing codes in four layers of coding layers in sequence;

[0029] In combination with the random strategy and the maximum processing time optimization strategy, the corresponding cases of each type of box and different workpieces corresponding to each process, and the corresponding cases of each process and different production machines corresponding to each process are sequentially divided into corresponding process codes and station codes in four encoding layers.

[0030] The box type code, the packing code, the process code and the station code are decoded to generate a plurality of groups of collaborative scheduling schemes.

[0031] In some embodiments of the first aspect, in combination with the greedy and random strategies, the corresponding cases of each type of box and the number of different workpieces corresponding to each consignee address corresponding to each type of box are sequentially divided into corresponding box type codes and packing codes in four encoding layers, including:

[0032] For each consignee address, the greedy strategy is used to generate an initial packing code and a box type code according to the principle that the use priority of the box type is that the large capacity is higher than the small capacity, and the used box is combined into the target group with the most workpiece types;

[0033] The initial packing code and the box type code are converted into the packing code and the box type code by randomly adjusting the packing order between each consignee address.

[0034] In combination with the random strategy and the maximum processing time optimization strategy, the corresponding cases of each type of box and different workpieces corresponding to each process, and the corresponding cases of each process and different production machines corresponding to each process are sequentially divided into corresponding process codes and station codes in four encoding layers, including:

[0035] First, randomly adapt different production machines to each process, and then generate process codes and station codes according to the principle that the shortest machine in the different production machines is adapted to each process, and the corresponding production machine selected in the previous process is not adapted by the subsequent process.

[0036] In some embodiments of the first aspect, the strategy parameter includes a control parameter, and step S4 of updating the strategy parameter includes:

[0037] If the population dispersion of the plurality of alternative collaborative scheduling schemes is higher than the preset dispersion index, the decay step of the control parameter is increased, if it is less than the preset dispersion index, the decay step of the control parameter is decreased, and if it is equal to the preset dispersion index, the decay step of the control parameter is kept unchanged, wherein the dynamic adjustment index of the control parameter is linearly positively correlated with the preset dispersion index;

[0038] Step S5: adjusting each alternative collaborative scheduling scheme based on a neighborhood search method, including:

[0039] In the process coding, two processes are randomly selected and exchanged, in the machine coding, a production machine is randomly replaced by a machine with the shortest processing time within the adaptable range, and / or an exchange workpiece operation is performed in the bin coding; wherein the exchange workpiece operation includes exchanging all workpieces in two bins from different types of bins or exchanging target workpieces in two bins with the same consignee address;

[0040] The hierarchical storage Archive population includes a core layer and a balance layer, and step S7: based on the fitness value of each candidate collaborative scheduling scheme, each group of collaborative scheduling scheme groups is non-dominantly sorted to update the hierarchical storage Archive population, including:

[0041] According to the fitness value of each candidate collaborative scheduling scheme, a new candidate collaborative scheduling scheme is determined to be exchanged between different collaborative scheduling scheme groups;

[0042] If the solution dominance relationship of the new candidate collaborative scheduling scheme and the candidate collaborative scheduling scheme in the core layer is the same, the similarity between the new candidate collaborative scheduling scheme and the candidate collaborative scheduling scheme in the balance layer is determined based on the coding distance calculation;

[0043] The candidate collaborative scheduling scheme with the highest similarity to the new candidate collaborative scheduling scheme in the balance layer is replaced to update the hierarchical storage Archive population.

[0044] In some embodiments of the first aspect, before obtaining the information matrix and the delivery time of the workpieces of each consignee address, the method further comprises:

[0045] Obtaining a plurality of order information, bin type information and production workshop information;

[0046] Constructing an order demand matrix according to the workpiece type, workpiece quantity and consignee address information in the plurality of order information;

[0047] Determining the delivery time of the workpieces of each consignee address in the plurality of order information;

[0048] Constructing a bin capacity matrix according to the bin type information;

[0049] According to the production workshop information, obtaining a workpiece processing time table, and constructing a processing time matrix;

[0050] According to the production workshop information, obtaining a workpiece occupied capacity table, and constructing a workpiece occupied capacity matrix;

[0051] The preset constraint conditions include:

[0052] 1) Each process of any workpiece in different workpieces is completed on only one production machine;

[0053] 4) for each process, the start time of the next process is greater than or equal to the start time of the previous process plus the processing time of the previous process;

[0054] For each type of box, the total capacity of the workpieces in each box does not exceed the rated capacity of the box.

[0055] The second aspect of the present application provides a customized production and boxing collaborative scheduling system, comprising: a processor and a memory; the memory is coupled with the processor, and the memory is used to store computer program codes; the processor invokes the computer program codes to make the system execute the method of the first aspect.

[0056] The third aspect of the present application provides a computer readable storage medium, and the computer readable storage medium stores a computer program; when the computer program is executed by a processor, the method of the first aspect is realized.

[0057] The fourth aspect of the present application provides a computer product, comprising a computer program; when the computer program is executed by a processor, the method of the first aspect is realized.

[0058] It can be understood that the customized production and boxing collaborative scheduling method, system, storage medium and product of the present application generate strategy parameters and multiple collaborative scheduling schemes according to information matrices including order demand matrix, box capacity matrix, processing time matrix and workpiece occupied capacity matrix, and the delivery period of workpieces of each delivery address, through the production and boxing collaborative scheduling model and by using the problem-specific knowledge method. Multiple collaborative factors of the two links of customized production and workpiece boxing are considered at the same time, the processing time can be accurately calculated, the production machines are reasonably distributed and the process sequence is arranged, the waiting time of the production machines and the boxing machines and the process connection time are reduced, so as to effectively shorten the production cycle, improve the production efficiency and enable the delivery to be made in stages, and improve the customer satisfaction.

[0059] At the same time, the occupied capacity of the workpiece, the type and capacity limit of the box, and each delivery address corresponding to the order information and other factors are fully considered, a scientific boxing scheme is adopted, the space of the box is maximally utilized, the waste of the box is reduced, and the boxing rate is significantly improved. The target collaborative scheduling scheme obtained based on the production and boxing collaborative scheduling model can take into account the market demand of high production efficiency and high boxing efficiency of customized orders. BRIEF DESCRIPTION OF DRAWINGS

[0060] The drawings incorporated into the specification and forming a part of the specification, show embodiments consistent with the present application, and together with the specification, serve to explain the principles of the present application.

[0061] Figure 1 A flowchart of the customized production and boxing collaborative scheduling method provided by the embodiments of the present application is shown in the figure.

[0062] Figure 2 An application scenario diagram of the customized production and boxing collaborative scheduling method provided by the embodiment of the present application is shown in the following figure.

[0063] Figure 3 Another application scenario diagram of the customized production and boxing collaborative scheduling method provided by the embodiment of the present application is shown in the following figure.

[0064] Figure 4 Still another application scenario diagram of the customized production and boxing collaborative scheduling method provided by the embodiment of the present application is shown in the following figure.

[0065] Figure 5 Still another application scenario diagram of the customized production and boxing collaborative scheduling method provided by the embodiment of the present application is shown in the following figure.

[0066] Figure 6 A scheduling Gantt chart of the scheme 1 of the customized production and boxing collaborative scheduling method of the present application is shown in the following figure.

[0067] Figure 7 A scheduling Gantt chart of the scheme 2 of the customized production and boxing collaborative scheduling method of the present application is shown in the following figure.

[0068] Figure 8 A structure diagram of the customized production and boxing collaborative scheduling system provided by the embodiment of the present application is shown in the following figure.

[0069] The specific embodiments of the present application have been shown in the above figures, and will be described in more detail hereinafter. These figures and the written description are not intended to limit the scope of the present application concept in any way, but to illustrate the present application concept to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0070] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The same or similar components are denoted by the same reference numerals throughout the drawings and the following description, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application.

[0071] The terms "first", "second", and the like in the present application are used only for the purpose of description, and should not be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated.

[0072] The technical solutions of the present application and how the technical solutions of the present application solve the technical problems will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0073] Please refer to Figure 1 , Figure 1A flowchart of a customized production and boxing collaborative scheduling method is provided in the present application. The execution subject of the scheduling method can be a customized production and boxing collaborative scheduling system, as shown in Figure 1 The scheduling method can include the following steps:

[0074] Step S1: Obtain the information matrix and the delivery period of the workpieces of each delivery address.

[0075] The information matrix includes an order demand matrix, a box capacity matrix, a processing time matrix, and a workpiece occupied capacity matrix.

[0076] Specifically, before performing step S110, obtain multiple order information, box type information, and production workshop information. Then, construct an order demand matrix according to the workpiece type, workpiece quantity, and delivery address information in the multiple order information The order demand matrix is used to represent the quantity of different workpieces required by each delivery address, where p is the workpiece type, q is the delivery address, m ij represents the quantity of workpiece i of delivery address j.

[0077] and determine the delivery period of the workpieces of each delivery address in the multiple order information where q represents the delivery address, and d j represents the delivery period of delivery address j.

[0078] Construct a box capacity matrix according to the box type information The box capacity matrix is used to represent the capacity of each type of box, where n represents the box type, and b i represents the capacity of the i-th type of box. According to the production workshop information, obtain the process processing time table, and construct a processing time matrix The processing time matrix is used to represent the processing time of each process of different workpieces on different machines, where represents the processing time of the j-th process of workpiece i on production machine k.

[0079] According to the production workshop information, obtain the workpiece occupied capacity table, and construct a workpiece occupied capacity matrix The workpiece occupied capacity matrix is used to represent the occupied capacity of different workpieces, p represents the workpiece type, and c i represents the occupied capacity of workpiece i .

[0080] In one application scenario, the multiple order information is shown in Table 1:

[0081] Table 1 Order information table

[0082]

[0083] In this application scenario, the order demand matrix is:

[0084]

[0085] According to the order table information, determine the delivery date of each delivery address. The expression is:

[0086]

[0087] The capacity table of box types is shown in Table 2.

[0088] Table 2 Box type capacity table

[0089]

[0090] In this application scenario, the box capacity matrix is:

[0091]

[0092] The workpiece occupied capacity table is shown in Table 3.

[0093] Table 3 Workpiece Occupancy Capacity

[0094]

[0095] Then in the application scenario, the workpiece capacity matrix is:

[0096]

[0097] The process processing schedule is shown in Table 4:

[0098] Table 4 Processing schedule

[0099]

[0100] In this application scenario, the processing time matrix is:

[0101]

[0102] Step S2: Using the problem-specific knowledge method, generate strategy parameters and multiple groups of collaborative scheduling solutions based on the information matrix and the delivery dates of the workpieces at each delivery address. Each group of collaborative scheduling solutions includes multiple alternative scheduling solutions for the same number of boxes.

[0103] For example, Figures 2 to 4 The individuals shown are coded A and C. Individuals A and B with the same number of boxes (both 3 boxes) are divided into one group, and C into another group.

[0104] It can be understood that, since the length of the process code and the machine code depends on the number of boxes, grouping individuals with the same number of boxes is adopted to facilitate subsequent population evolution operations. The grouping evolution mechanism is introduced to solve the problem of different lengths of the code of different packing schemes, and other algorithms such as genetic algorithm need to design a complex repair mechanism, which is easy to produce invalid solutions.

[0105] The strategy parameters include a control parameter a, a direction weight A, and a disturbance factor C.

[0106] The problem-specific knowledge method includes a greedy strategy, a random strategy, and a maximum processing time optimization strategy.

[0107] In an embodiment, when the problem-specific knowledge method is used, preset assumption conditions are met at the same time. The preset assumption conditions include:

[0108] (1) The process flow and processing time of different workpieces are known and determined.

[0109] (2) The processing time required by each process does not change with the production order or other factors.

[0110] (3) For the same type of multiple quantities of workpieces, the processing time calculation follows the simple accumulation principle. For example, if there are workpieces 1, each workpiece 1 requires time to complete process j (j represents the process number), and the process quantity of workpiece 1 is , then the total process processing time of each workpiece 1 is . This is based on the fact that each workpiece 1 needs to go through all its own processes in turn, and the processing time of each workpiece is accumulated to obtain the total processing time of all workpieces of this type.

[0111] (4) The conversion time between processes can be ignored. When a production machine completes a process of a workpiece and switches to the next process, or when it switches from processing one workpiece to processing another workpiece, the preparation time, equipment adjustment time, etc. are all considered as zero.

[0112] Specifically, the generation method of the alternative coordination scheduling scheme includes: under the condition of meeting the preset assumption conditions, according to the information matrix and the delivery period of the workpieces of each delivery address, combining the greedy and random strategies, the corresponding type of each type of box and the corresponding situation of the number of different workpieces of each delivery address corresponding to each type of box are divided into the corresponding box type code and packing code in four coding layers in turn.

[0113] Then, by combining the random strategy and the maximum processing time optimization strategy, the corresponding processes of different workpieces in various types of boxes and the corresponding situations of different production machines are divided into corresponding process codes in four coding layers (the process code length is la (la = the number of all processes * the number of boxes, indicating the execution order of each process)) and workstation codes (workstation codes lb (lb = the number of all processes * the number of boxes), indicating the allocation of processes to production machines.

[0114] Finally, the box type code, packing code, process code, and workstation code are decoded to generate multiple sets of collaborative scheduling plans. Each alternative scheduling plan in each set of collaborative scheduling plans includes information such as the total number of addresses at each delivery address, the total capacity of the boxes corresponding to each delivery address and the actual capacity occupied by the workpieces in the boxes, the delivery date of the workpieces required at each delivery address when the last workpiece at each delivery address is completed, the demand for different workpieces for each order in different time periods, the time required for different workpieces to complete each process, and whether each process of different workpieces is processed on different production machines.

[0115] For example, the first layer is the box type code, and the digital code 1, 2, 3 represents the box type. The second layer is the packing code. For example, 2.11, 2.12, 2.13, each three digits of the code constitutes a box. The first digit before the decimal point represents the address, and the digit after the decimal point represents the number of workpieces. 2.11, 2.12, 2.13 means that 11 workpieces 1, 12 workpieces 2, and 13 workpieces 3 at address 2 constitute a box. The third layer is the process code with a length of l a (l a = total number of processes * number of boxes), the tens digit of the digital code represents the box number, the ones digit represents the workpiece number, and the number of times it appears repeatedly represents the process of the workpiece in the box. In this example, the first occurrence of the first digit 11 in the digital code indicates the first process of workpiece 1 in the first box, and the second occurrence of the fourth digit 11 indicates the second process of workpiece 1 in the first box; the fourth layer is the station code with a length of l b (l b = total process quantity * box quantity), digital codes 1, 2, 3, indicating that the process is assigned to machines 1, 2, 3.

[0116] like Figures 2 to 4 As shown, each alternative collaborative scheduling scheme (code A, code B or code C) includes the first-level box type code, the second-level packing code, the third-level process code and the fourth-level workstation code.

[0117] It can be understood that this embodiment adopts a dedicated coding structure of box type-packing-process-station designed specifically for the production-packing collaborative problem, directly mapping the actual decision variables to break the traditional production and packing separation decision-making model and achieve collaborative optimization of the two.

[0118] Step S3: Non-dominated sorting is performed on each group of collaborative scheduling schemes based on the multiple objective values obtained by inputting each alternative scheduling scheme into the production and boxing collaborative scheduling model, and a hierarchical storage Archive population is established.

[0119] Specifically, the solving method of the production and boxing collaborative scheduling model includes a grey wolf optimization algorithm. The grey wolf population in the grey wolf optimization algorithm is regarded as multiple alternative collaborative scheduling schemes, and the grey wolf individuals of the grey wolf population are regarded as each alternative collaborative scheduling scheme. Multiple initial optimization parameters corresponding to the alternative collaborative scheduling schemes are set, and the initial optimization parameters include the number of alternative collaborative scheduling schemes, the maximum number of iterations, the unit cost of early completion, and the unit cost of delay.

[0120] The production and boxing collaborative scheduling model maximizes the average boxing volume utilization rate, minimizes the maximum completion time of multiple order information, and minimizes the early / delay cost, and is constructed in combination with the multi-objective optimization theory. The target collaborative scheduling scheme indicates that different production machines process different workpieces in different types of boxes in different processes, and that the boxing machine packs different workpieces into corresponding different types of boxes.

[0121] Step S4: The strategy parameters are updated, and the collaborative scheduling representative scheme is selected from the hierarchical storage Archive population by the roulette method. The fitness value of the collaborative scheduling representative scheme is adjusted to adjust the other alternative collaborative scheduling schemes.

[0122] Specifically, the update strategy parameters include dynamically adjusting the control parameters, the update direction weight, and the disturbance factor. The collaborative scheduling representative scheme is regarded as the head wolf of the grey wolf population 、 、 The production and boxing collaborative scheduling model based on the collaborative scheduling representative scheme obtains the fitness value, and the influence weight of the collaborative scheduling representative scheme on the update of the other alternative collaborative scheduling schemes in the collaborative scheduling scheme group is dynamically adjusted according to the fitness value to adjust the other alternative collaborative scheduling schemes.

[0123] It can be understood that the production and boxing collaborative scheduling model is the objective function, and the corresponding fitness value can be obtained according to the objective function. The influence weight of the collaborative scheduling representative scheme on the update of the other alternative collaborative scheduling schemes in the collaborative scheduling scheme group is dynamically adjusted according to the fitness value to adjust the other alternative collaborative scheduling schemes. The adjustment formula is wherein, represents the other alternative collaborative scheduling scheme, represents the collaborative scheduling representative scheme represents the collaborative scheduling representative scheme 、 、 the influence weight of the collaborative scheduling representative scheme, and the expression is . wherein: , is an initial coefficient, T max is a maximum iteration number, is a decay index, represents a collaborative scheduling representative scheme under the ith objective function (i.e. under the ith production and binning collaborative scheduling model) , , fitness value.

[0124] Step S5: adjusting each candidate collaborative scheduling scheme based on a neighborhood search method.

[0125] Step S6: judging whether the adjusted candidate collaborative scheduling scheme satisfies a preset constraint condition, if not, executing step S5, and if yes, executing step S7.

[0126] wherein, the preset constraint condition comprises:

[0127] 1) a unique machine allocation constraint, each process of any workpiece in different workpieces is completed on and only on one production machine; i.e. , represents whether the process j of the workpiece i is processed on the production machine k, 1 if yes, and 0 otherwise.

[0128] 2) a process sequence constraint, for each process, the start time of the next process is greater than or equal to the start time of the previous process plus the processing time of the previous process; i.e. wherein, represents the start time of the j+1th process of the workpiece i, represents the start time of the jth process of the workpiece i, represents the processing time of the jth process of the workpiece i.

[0129] 3) a bin capacity constraint, for each type of bin, the total capacity of the workpieces loaded in each bin does not exceed the rated capacity of the bin. i.e. wherein p i represents the capacity occupied by the workpiece i, N yi represents the number of the workpiece i in the yth bin, and C n represents the rated capacity of the nth bin.

[0130] Step S7: non-dominant sorting each group of collaborative scheduling schemes based on the fitness value of each candidate collaborative scheduling scheme to update the hierarchical storage Archive population.

[0131] In an embodiment, the new candidate collaborative scheduling scheme exchanged between different groups of collaborative scheduling schemes is determined according to the fitness value of each candidate collaborative scheduling scheme.

[0132] Then, if the new candidate co-scheduling scheme has the same dominance relationship with the candidate co-scheduling scheme in the core layer, the similarity between the new candidate co-scheduling scheme and the candidate co-scheduling scheme in the balance layer is determined based on the encoding distance calculation.

[0133] It can be understood that the hierarchical storage Archive population is stored according to the target contribution degree: 1) the core layer: storing the optimal solution of each target (such as the optimal solution of the average container volume utilization rate, the shortest solution of the maximum completion time), 2) the balance layer: storing the multi-objective balanced solution. The steps of calculating the similarity based on the encoding distance include:

[0134] The encoding distance calculation formula is , wherein is the weight of each layer (the box type encoding layer =0.2, the packing encoding layer =0.3, the process encoding layer =0.3, and the workstation encoding layer =0.2), is the distance function of each layer.

[0135] The box type encoding distance function is , wherein is used to determine whether the two box encoding vectors X and Y are different at the first k position, if different, then =1, otherwise 0,

[0136] The packing encoding distance function is , and m is the number of same address workpieces.

[0137] The process encoding distance calculation method is the edit distance, and the specific function is .

[0138] The machine encoding distance function is .

[0139] Similarity: D=D1+D2+D3+D4.

[0140] Finally, the candidate co-scheduling scheme with the highest similarity to the new candidate co-scheduling scheme in the balance layer is replaced to update the hierarchical storage Archive population.

[0141] That is, in the balance layer, the Sim of the new solution and all old solutions is calculated, and the old solution with the maximum Sim (i.e., the minimum distance) is selected to replace. It can be understood that when the new solution (i.e., the new candidate collaborative scheduling scheme) is added, if the domination relationship with the core layer solution is the same, the solution with the highest similarity (i.e., the candidate collaborative scheduling scheme with the highest similarity) in the balance layer is replaced.

[0142] It can be understood that for the four-layer coding structure, the updating mechanism of the present application is better than other operations such as the cross mutation operation of the genetic algorithm in maintaining the effectiveness of the coding, and converges better in optimizing the volume utilization rate target, and can find a more compact packing scheme.

[0143] Step S8: Determine whether the number of Archive populations reaches the preset upper limit, if yes, eliminate the candidate coordination scheduling scheme with a congestion degree greater than the preset congestion degree, and if not, execute step S9.

[0144] Step S9: Determine whether the maximum number of iterations is reached, if not, re-execute step S4, and if yes, output the collaborative scheduling scheme group, and determine the target collaborative scheduling scheme from the multiple collaborative scheduling scheme groups based on the multiple target values of the current iteration.

[0145] In an embodiment, the average packing volume utilization rate, the maximum completion time, and the early / delayed cost are comprehensively analyzed, and in the multiple collaborative scheduling scheme groups, the candidate collaborative scheduling scheme with the maximum average packing volume utilization rate, the maximum completion time of the multiple order information, and the optimal early / delayed cost is selected as the target collaborative scheduling scheme.

[0146] It can be understood that in the above technical solution, the production and packing collaborative scheduling model is constructed by maximizing the average packing volume utilization rate, minimizing the maximum completion time of the multiple order information, and minimizing the early / delayed cost as the optimization target, combined with the multi-objective optimization theory, and the problem-specific knowledge method and the preset assumption condition are used to generate multiple candidate collaborative scheduling schemes of the customized production and material packing collaborative scheduling based on the information matrix and the delivery time of the workpieces of each delivery address, and the candidate scheduling scheme corresponding to the candidate collaborative scheduling scheme is generated. After the packing collaborative scheduling model of the candidate scheduling scheme is input for iterative calculation, the target collaborative scheduling scheme is determined from the multiple candidate collaborative scheduling schemes based on the multiple target values obtained in the current iteration, and the target collaborative scheduling scheme can meet the market demand of high production efficiency and high packing efficiency of customized orders.

[0147] Specifically, by producing a production and boxing collaborative scheduling model and using a problem-specific knowledge method, strategy parameters and multiple collaborative scheduling schemes are generated according to information matrices including an order demand matrix, a box capacity matrix, a processing time matrix, a workpiece occupied capacity matrix, and a delivery period of workpieces of each consignee address, while considering multiple collaborative factors of custom production and workpiece boxing, the processing time can be accurately calculated, the production machines and the process sequence can be reasonably allocated, and the waiting time of the production machines and the boxing machines and the process connection time can be reduced, thereby effectively shortening the production cycle, improving the production efficiency, and enabling the delivery to be made in stages, and improving the customer satisfaction.

[0148] Meanwhile, by fully considering the occupied capacity of the workpieces, the type and capacity limit of the boxes, and the consignee addresses corresponding to the order information, a scientific boxing scheme is adopted to maximize the use of the space of the boxes, reduce the waste of the boxes, and significantly improve the boxing rate. The target collaborative scheduling scheme obtained based on the production and boxing collaborative scheduling model can take into account the market demand for high production efficiency and high boxing efficiency of customized orders.

[0149] In some embodiments, in step S2, in combination with the greedy and random strategies, the corresponding cases of the type of each type of box and the number of different workpieces of each consignee address corresponding to each type of box are divided into corresponding box type codes and boxing codes in four layers of coding layers in turn, including:

[0150] First, for each consignee address, the greedy strategy is adopted to generate the initial boxing code and the box type code according to the principle that the use priority of the box type is that the large capacity is higher than the small capacity, and the target combination of the used boxes is the combination of the most workpiece types. For example, for address one, the largest capacity box is first tried to be used to pack as many combinations of workpieces 1, 2, and 3 as possible, and the remaining is packed in smaller boxes. Similarly, address two is processed, and after repeated multiple times, multiple different initial boxing codes and box type codes corresponding to the alternative collaborative scheduling schemes are obtained.

[0151] Then, the boxing order between the consignee addresses is randomly adjusted to convert the initial boxing code and the box type code into the boxing code and the box type code.

[0152] For example, the initial boxing order of address one first and address two second is adjusted to the boxing order of address two first and address one second. The boxing code and the box type code corresponding to the alternative collaborative scheduling schemes are obtained

[0153] It can be understood that in the present embodiment, the initial boxing code and the box type code are generated by the greedy strategy first, and then adjusted by the random strategy, so that the alternative collaborative scheduling schemes have diversity.

[0154] In some embodiments, in step S2, in combination with the random strategy and the maximum processing time optimization strategy, the corresponding cases of each process in each type of box and the corresponding cases of different production machines corresponding to each process are sequentially divided into corresponding process codes and station codes in four encoding layers, including: first, randomly adapting different production machines for each process, then according to the shortest processing time of each process respectively adapting different production machines, and in the principle that the corresponding production machine selected in the previous process is no longer adapted by the subsequent process, generating process codes and station codes.

[0155] It can be understood that in this embodiment, according to the type of machine required by the process, all machines suitable for the process are selected from the machine type. Then, based on the maximum processing time optimization, the machine with shorter processing time is selected to perform the processing task. When a process requiring processing is encountered subsequently, the machine is selected from the machines that have not been selected before, and the above-mentioned principle of selecting and using the machine with shorter processing time is continuously followed until all machines suitable for the process have been used once, which can ensure that each machine can participate in the processing flow of the corresponding process at the appropriate time.

[0156] In some embodiments, the plurality of target values includes average binning volume utilization, maximum completion time, and early / delayed cost. The production and binning collaborative scheduling model includes a binning average volume utilization determination model, a completion time determination model, and an early / delayed cost determination model. That is, in step S3, based on the plurality of target values obtained by inputting each alternative scheduling scheme into the production and binning collaborative scheduling model, the following steps are included:

[0157] Step a.1: determining the total number of addresses of each consignee address, the total capacity of the box corresponding to each consignee address, and the actual occupied capacity of the workpiece in the box in the alternative scheduling scheme;

[0158] Step a.2: determining the binning volume utilization of each consignee address according to the total capacity of the box and the actual occupied capacity of the workpiece in the box;

[0159] Step a.3: obtaining the average binning volume utilization according to the binning volume utilization of each consignee address and the total number of addresses;

[0160] It can be understood that after inputting the alternative scheduling scheme into the binning average volume utilization determination model, the calculation process of the average binning volume utilization can be obtained.

[0161] Exemplarily, the expression of the binning average volume utilization determination model is (1). Wherein, represents the average binning volume utilization, n represents the total number of addresses of each consignee address, r i represents the binning volume utilization of the i-th consignee address. represents the total capacity of all the boxes sent to the i-th consignee address, which is the sum of the box capacities (i.e., the total capacity of the boxes sent to each consignee address). For example, if m boxes are used to send to the consignee address, the box capacities are . represents the actual capacity of the workpieces in the box sent to the i-th consignee address.

[0162] Step a.4: determining the completion time of the last workpiece in each consignee address in the alternative scheduling scheme;

[0163] Step a.5: accumulating the last completion time to obtain the maximum completion time;

[0164] It can be understood that after the alternative scheduling scheme is input into the completion time determination model, the calculation process of minimizing the maximum completion time can be obtained.

[0165] Exemplarily, the expression of the completion time determination model is wherein, represents the minimum maximum completion time, represents the completion time of the last workpiece in the consignee address r.

[0166] Step a.6: obtaining the early completion time and the delay time of each consignee address according to the delivery period of the workpieces required by each consignee address in the alternative scheduling scheme and the completion time of the last workpiece in each consignee address.

[0167] Step a.7: obtaining the early / delay cost according to the early completion time and the delay time, and the preset early completion unit cost and delay unit cost.

[0168] It can be understood that after the alternative scheduling scheme is input into the early / delay cost determination model, the calculation process of the early / delay cost can be obtained.

[0169] Exemplarily, the expression of the early / delay cost determination model is: wherein, represents the early / delay cost, and are the preset early completion unit cost and delay unit cost, respectively, is the delivery period of the workpieces required by the consignee address r, r is the early completion time of the consignee address r, is the delay time of the consignee address r. r r represents the completion time of the last workpiece in the consignee address r.

[0170] In one application scenario in the embodiment, the maximum completion time and the early / delay cost can be generated Figures 2 to 4 ​​​The three alternative scheduling schemes, encoding A, encoding B and encoding C, correspond to three target values of average container volume utilization, maximum completion time and early / late cost.

[0171] Specifically, the average container volume utilization of individual encoding A is 74%, the maximum completion time is 1760, and the early / late cost is 255; the average container volume utilization of individual encoding B is 81.8%, the maximum completion time is 1800, and the early / late cost is 567; and the average container volume utilization of individual encoding C is 79.5%, the minimum completion time is 1790, and the early / late cost is 426. It can be seen that encoding A is more advantageous in maximum completion time and early / late cost, and encoding B is more advantageous in average container volume utilization.

[0172] In some embodiments, the dynamic adjustment of the control parameter a in step S4 includes:

[0173] If the population dispersion of the plurality of alternative collaborative scheduling schemes is higher than the preset dispersion index, the decay step of the control parameter is increased to accelerate the contraction.

[0174] If it is less than the preset dispersion index, the decay step of the control parameter is reduced to prolong the exploration; if it is equal to the preset dispersion index, the decay step of the control parameter is kept unchanged.

[0175] The dynamic adjustment index of the control parameter is linearly positively correlated with the preset dispersion index.

[0176] Exemplarily, the adjustment formula of the control parameter a is .

[0177] wherein is the dynamic adjustment index, which is adaptively calculated according to the preset dispersion index, such as , α is a preset weight factor, is the preset dispersion index.

[0178] It can be understood that in the traditional grey wolf optimization algorithm, the control parameter a (which determines the contraction / expansion speed of surrounding prey) is usually decayed according to a fixed formula. Such a fixed strategy may lead to early convergence too fast: if a decays too fast, the algorithm falls into local optimum too early; late search stagnation: if a decays too slowly, the algorithm is difficult to search finely. In the present embodiment, the decay rate of a is adjusted in real time according to the population diversity or the convergence state, which can improve the fineness of the search.

[0179] In some embodiments, the neighborhood search method includes local search, and step S5: adjusting each alternative collaborative scheduling scheme based on the neighborhood search method includes:

[0180] In process coding, two processes are randomly selected for interchange; in machine coding, a production machine is randomly replaced with the machine with the shortest processing time within the adaptable range and / or workpiece interchange operations are performed in packaging coding.

[0181] The workpiece swapping operation includes swapping all workpieces in two boxes from various boxes or swapping target workpieces in two boxes with the same delivery address. Figure 5 As shown, all artifacts in the box coded as 3 are swapped with all artifacts in the box coded as 2.

[0182] The following is a comparative example compared with the method of the present application to discuss the superiority of the method of the present application.

[0183] Specifically, the target collaborative scheduling scheme generated by applying the customized production and packaging collaborative scheduling method of this application is Scheme 1.

[0184] Scheme 2 is a scheduling scheme generated without considering the coordinated scheduling of customized production and packing (i.e., using other algorithms).

[0185] Table 5 compares the content of Scheme 1 and Scheme 2, including the processing sequence of the processes, the corresponding production machines, and the packaging plan. Table 6 shows the corresponding target values ​​of the two schemes.

[0186] Table 5 Comparison of the contents of Scheme 1 and Scheme 2

[0187]

[0188] In solution 1 The jth process of workpiece i in the nth box is processed on machine k. In the packing plan, (11-12-13)->3 means that 11 pieces of workpiece 1, 12 pieces of workpiece 2, and 13 pieces of workpiece 3 are packed in box type 3.

[0189] In solution 2 It means that the j-th process of workpiece i is processed on machine k, and the packing plan is the same as above.

[0190] Table 6 Target value comparison

[0191]

[0192] Combined with Table 6, Figure 6 Scheme 1 Scheduling Gantt Chart and Figure 7 The Gantt chart of Scheme 2 shows that Scheme 1 has more advantages in terms of maximum completion time and minimum advance / delay functions. In the phased delivery scenario, Scheme 1 is the best and preferred option.

[0193] It can be understood that the customized production and boxing collaborative scheduling method of the application, the special coding structure of the box type-boxing-process-station is designed for the production-boxing collaborative problem, which directly maps the actual decision variables to break the traditional production and boxing separate decision mode, and realizes the collaborative optimization of the two. For the four-layer coding structure, the updating mechanism of the method is more effective than the cross variation operation and other operations of the genetic algorithm, and the convergence effect is better when optimizing the volume utilization rate target, and a more compact boxing scheme can be found. The grouping evolution mechanism is introduced to solve the problem of different coding lengths of different boxing schemes. Genetic algorithm and other algorithms need to design a complex repair mechanism, which is easy to produce invalid solutions.

[0194] Figure 8 The structure diagram of the customized production and boxing collaborative scheduling system provided by the application is shown in FIG. 1. Figure 8 As shown in FIG. 1, the customized production and boxing collaborative scheduling system 10 comprises:

[0195] a processor 11, a memory 12 and a bus 13;

[0196] The memory 12 is used to store the computer program code of the processor 11;

[0197] The processor 11 is configured to execute the technical solutions of the customized production and boxing collaborative scheduling method in any of the preceding method embodiments by executing the computer program code.

[0198] The customized production and boxing collaborative scheduling system 10 is used to execute the technical solutions provided in any of the preceding method embodiments, and the implementation principles and technical effects are similar, which will not be described here.

[0199] The application also provides a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program is executed by the processor to implement the customized production and boxing collaborative scheduling method as described above.

[0200] The application also provides a computer product comprising a computer program, which is executed by the processor to implement the customized production and boxing collaborative scheduling method as described above.

[0201] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware. The aforementioned program can be stored in a computer readable storage medium. The program is executed to perform the steps of the above-mentioned method embodiments; and the aforementioned storage medium includes ROM, RAM, magnetic disc or optical disc and various storage medium capable of storing program code.

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for coordinated scheduling of customized production and packaging, characterized in that: Methods include: Step S1: Obtaining an information matrix and the delivery date of workpieces of each delivery address, wherein the information matrix includes an order demand matrix, a box capacity matrix, a processing time matrix, and a workpiece occupied capacity matrix; Step S2: Based on the information matrix and the delivery dates of the workpieces at each delivery address, a greedy strategy is used for each delivery address. The priority of using box types is to prioritize large capacity over small capacity, and the box type with the largest number of workpiece types is used as the target combination. This strategy generates the initial packing code and box type code. Randomly adjust the packing order between each delivery address, and convert the initial packing code and box type code into a packing code and box type code; First, randomly adapt different production machines to each process, and then adapt each process to the machine with the shortest processing time among different production machines. The production machine selected in the previous process will no longer be adapted by the subsequent process. Generate process codes and workstation codes; Decode the box type code, packing code, process code, and workstation code to generate multiple sets of collaborative scheduling plans and generate strategy parameters, where each set of collaborative scheduling plans includes multiple alternative scheduling plans for the same number of boxes; Step S3: Based on the multiple target values ​​obtained by inputting each alternative scheduling plan into the production and packaging collaborative scheduling model, each group of collaborative scheduling plans is non-dominated sorted to establish a hierarchically stored Archive population; Step S4: Update the strategy parameters, and select the collaborative scheduling representative scheme from the hierarchically stored Archive population through the roulette wheel method, and adjust other alternative collaborative scheduling schemes according to the fitness value corresponding to the collaborative scheduling representative scheme; Step S5: adjusting each candidate coordinated scheduling solution based on a neighborhood search method; Step S6: Determine whether the adjusted alternative coordinated scheduling scheme meets the preset constraint conditions. If not, execute step S5; if yes, execute step S7. Step S7: performing non-dominated sorting on each group of collaborative scheduling solutions based on the fitness value of each alternative collaborative scheduling solution to update the hierarchically stored Archive population; Step S8: Determine whether the number of Archive populations reaches a preset upper limit. If yes, remove the alternative coordination scheduling schemes whose congestion is greater than the preset congestion. Otherwise, execute step S9. Step S9: Determine whether the maximum number of iterations has been reached. Otherwise, re-execute step S4. If so, output a collaborative scheduling solution group. Based on multiple target values ​​of the current iteration, determine a target collaborative scheduling solution from multiple collaborative scheduling solution groups.

2. The method according to claim 1, characterized in that The production and packing collaborative scheduling model is constructed by combining multi-objective optimization theory with the optimization objectives of maximizing the average packing volume utilization, minimizing the maximum completion time of multiple order information, and minimizing the lead / delay cost.

3. The method according to claim 1 or 2, characterized in that The multiple target values ​​include average container volume utilization, maximum completion time, and lead / delay cost. Step S3: inputting the multiple target values ​​obtained by the production and container collaborative scheduling model based on each alternative scheduling solution includes: Determine the total number of addresses for each delivery address in the alternative scheduling plan, the total capacity of the boxes corresponding to each delivery address, and the actual capacity occupied by the workpieces in the boxes; Determining the packing volume utilization rate of each delivery address based on the total capacity of the box and the actual capacity occupied by the workpieces in the box; Obtaining an average packing volume utilization rate based on the packing volume utilization rate of each delivery address and the total number of addresses; and Determine the completion time of the last workpiece in each delivery address in the alternative scheduling plan; Accumulating the final completion times to obtain a maximum completion time; and According to the delivery date of the workpieces required by each delivery address in the alternative scheduling plan and the completion time of the last workpiece in each delivery address, the early completion time and the delayed completion time of each delivery address are obtained; The early completion time and the delayed completion time, as well as the preset early completion unit cost and delayed unit cost, are used to obtain the early completion / delay cost.

4. The method according to claim 1, wherein The strategy parameters include control parameters. Step S4: updating the strategy parameters includes: If the population dispersion of multiple alternative coordinated scheduling schemes is higher than the preset dispersion index, the attenuation step size of the control parameter is increased. If it is lower than the preset dispersion index, the attenuation step size of the control parameter is decreased. If it is equal to the preset dispersion index, the attenuation step size of the control parameter is kept unchanged. The dynamic adjustment index of the control parameter is linearly positively correlated with the preset dispersion index. The step S5: adjusting each candidate coordinated scheduling solution based on the neighborhood search method, includes: In process coding, two processes are randomly selected for swapping; in machine coding, a production machine is randomly replaced with the one with the shortest processing time within the adaptable range; and / or in box coding, a workpiece swap operation is performed. The workpiece swap operation includes swapping all workpieces in two boxes of different types or swapping target workpieces in two boxes with the same delivery address. The hierarchically stored Archive population includes a core layer and a balance layer. Step S7: performing non-dominated sorting on each group of collaborative scheduling schemes based on the fitness value of each alternative collaborative scheduling scheme to update the hierarchically stored Archive population includes: Determine a new alternative coordinated scheduling scheme to be exchanged among different coordinated scheduling scheme groups according to the fitness value of each alternative coordinated scheduling scheme; If the new candidate co-scheduling scheme has the same solution dominance relationship as the candidate co-scheduling scheme in the core layer, then the similarity between the new candidate co-scheduling scheme and the candidate co-scheduling scheme in the balancing layer is determined based on the coding distance calculation; Replace the alternative collaborative scheduling scheme with the highest similarity to the new alternative collaborative scheduling scheme in the balancing layer to update the archive population of the hierarchical storage.

5. The method according to claim 1, wherein Before obtaining the information matrix and the delivery date of the workpieces at each delivery address, the method further includes: Get multiple order information, box type information and production workshop information; Building an order demand matrix based on the workpiece type, workpiece quantity and delivery address information in the plurality of order information; Determining the delivery date of the workpiece for each delivery address in the plurality of order information; Construct a box capacity matrix based on box type information; According to the production workshop information, obtain the process processing schedule and build the processing time matrix; According to the production workshop information, obtain the workpiece occupancy capacity table and build the workpiece occupancy capacity matrix; The preset constraints include: 1) Each process of any workpiece among different workpieces is completed on one and only one production machine; 2) For each process, the start time of the next process is greater than or equal to the start time of the previous process plus the processing time of the previous process; 3) For all types of boxes, the total capacity of workpieces in each box shall not exceed the rated capacity of the box.

6. A customized production and packaging collaborative scheduling system, characterized by: include: processor and memory; The memory is coupled to the processor, and the memory is used to store computer program code. The processor calls the computer program code to enable the system to perform the method according to any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

8. A computer product comprising a computer program, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images