A clothing hanging production scheduling method and system considering transportation distance

Abstract

The present invention discloses a method and system for scheduling clothing hanging production considering transportation distance. The scheduling method comprises the following steps: obtaining and processing information of a clothing hanging production workshop to obtain a worker proficiency coefficient matrix and a workstation distance matrix of the hanging workshop, establishing a mathematical model for clothing hanging production scheduling considering transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints, and introducing an improved non-dominated sorting genetic algorithm for solving the problem, so as to obtain a solution for clothing hanging production scheduling. The multi-objective optimization algorithm adopted by the present invention, in combination with the actual layout and production situation of the clothing hanging workshop, fully considers the optimization of the running route of the hanging line hanger for clothing hanging production scheduling, introduces initialization strategies such as path optimization, and an improved cross-mutation operation, has a small number of optimization iterations, can effectively avoid falling into a local optimal solution, has a strong multi-objective optimization capability, and has a high quality of solution.

Description

Technical Field

[0001] The present invention relates to the technical field of job shop scheduling, and in particular to a method and system for scheduling clothing hanging production taking transportation distance into consideration. Background Art

[0002] With the digital and intelligent transformation and upgrading of manufacturing enterprises, more and more clothing manufacturers are using clothing hanging systems to replace the original bundling and push-frame production. Clothing hanging is a type of flexible manufacturing system. The flexible transportation of hanging lines can greatly reduce the auxiliary operation time of material handling and the backlog of work-in-progress and cut pieces at the workstation, improve the punctuality of production, and have real-time adjustment capabilities through program control, which can quickly adapt to the flexible conversion of different varieties and batches.

[0003] However, the existing clothing hanging workshops completely rely on the production experience of the workshop team leader for production scheduling. When market demand changes rapidly and orders are frequently updated, the manual scheduling scheme often lacks scientificity and rationality, leading to problems such as waste of workshop production resources and delayed delivery. Studies and statistics show that in the process of clothing hanging production, the transportation time of hanging accounts for 30%-40% of the entire production time, and the existing clothing hanging scheduling often does not consider the optimization of the hanging transportation path. Under the scheduling scheme that does not consider the transportation distance, the increase in transportation time will directly extend the production cycle, thereby reducing the efficiency of the overall production line. The workstation load is also closely related to the transportation distance. If the transportation distance is too long, some workstations may have too much waiting time, resulting in uneven workstation load. And long-distance transportation also increases certain energy consumption and costs. Therefore, it is very necessary to consider the optimization of transportation distance in the optimization of clothing hanging production scheduling. Summary of the invention

[0004] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description and other drawings of the description.

[0005] The purpose of the present invention is to overcome the above-mentioned shortcomings and provide a method and system for scheduling clothing hanging production taking into account transportation distance.

[0006] The technical solution adopted by the present invention to solve the technical problem is: a clothing hanging production scheduling method considering the transportation distance, comprising the following steps:

[0007] Step S1, obtaining information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production log, and clothing hanging workshop layout;

[0008] Step S2, processing the information of the garment hanging production workshop to obtain the worker proficiency coefficient matrix and the workstation distance matrix of the hanging workshop;

[0009] Step S3, establishing a mathematical model for clothing hanging production scheduling considering transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints;

[0010] Step S4, introducing an improved non-dominated sorting genetic algorithm to solve the established mathematical model of clothing hanging production scheduling, so as to obtain a solution for clothing hanging production scheduling.

[0011] In some embodiments, in step S1, obtaining information of a clothing hanging production workshop includes:

[0012] Collect and pre-process order information, process information and workstation information based on garment production information;

[0013] Collect clothing hanging hanger flow history data based on clothing hanging production logs;

[0014] The distance between workstations and the speed of hanger movement are collected according to the layout of the clothing hanging workshop.

[0015] In some embodiments, pre-processing order information, process information and workstation information is specifically as follows:

[0016] The processing time matrix is ​​established based on the standard processing time of the order process in the obtained process information. Its expression is:

[0017]

[0018] In the above formula, n is the number of orders, m for n The maximum number of processes in an order;

[0019] Create a workstation machine type array based on the workstation number and machine type in the workstation information. The expression is:

[0020]

[0021] In the above formula, k is the largest workstation number, t is the largest machine type number, w ij The value is 0 or 1. w ij =1 indicates workstation i The machine type is j , a workstation can only use machines of one machine type, i.e. ;

[0022] According to the maximum hanging amount in the obtained workstation information, the workstation maximum hanging amount array is established, and its expression is:

[0023]

[0024] According to the processing matching of the obtained order process and workstation, a process-workstation matching matrix is ​​established, and its expression is:

[0025]

[0026] In the above formula, m i For orders O i The number of processes, c ijk Indicates order O i No. j Is the first process at the workstation? k If it can be processed, it is 1, otherwise it is 0.

[0027] In some embodiments, in step S2, processing the information of the garment hanging production workshop includes:

[0028] According to the hanger flow history data obtained from the clothing hanging production workshop information, the proficiency coefficients of the designated workers at different workstations in the processing procedures are calculated, and the worker proficiency coefficient matrix is ​​statistically established. The expression is:

[0029]

[0030] In the above formula, p ijk Indicates workstation k Processing Orders O i No. j The proficiency coefficient of a process is expressed as:

[0031] ,

[0032] In the above formula, α is the efficiency weight factor, β is the quality weight factor, satisfying α + β =1, the enterprise can determine it according to production strategy and market demand, and make corresponding adjustments according to the efficiency and quality requirements of different customers and orders; t ijk Indicates workstation k Processing Orders O i No. jThe actual processing time of a process is calculated by collecting historical data on the flow of hangers to calculate the time when hangers enter and leave the workstation during the process. k time difference; t ij Indicates processing order O i No. j The standard processing time of each process; N i For orders O i The total production quantity of N r For workstation k The amount of rework required to process this operation; t ijk The calculation formula is:

[0033] ,

[0034] In the above formula, t 进入 , t 离开 Workstations k Processing Orders O i No. j Each hanger enters and leaves the workstation during the process k time.

[0035] According to the distance between workstations in the obtained clothing hanging production workshop information, the workstation distance matrix is ​​established using the adjacency matrix method, and the expression is:

[0036] ,

[0037] In the above formula, d is the distance between adjacent workstations on the same side, and is the distance between the adjacent stations of the U-shaped bend of clothing hanging. The above matrix assumes that there are 4 stations in the hanging workshop, stations 1, 2, 3, and 4 are on the same side, so the distance from station 1 to station 3 is , and because the transportation direction is unidirectional, station 2 needs to go around to stations 3 and 4 to reach station 1, and the distance is .

[0038] In some embodiments, in step S3, optimizing the objective function includes:

[0039] The objective function of minimizing the maximum completion time is f 1, whose expression is:

[0040] ,

[0041] In the above formula, C max Indicates the maximum completion time for processing all orders. C i Indicates order i The completion time, n Indicates the total number of orders, R Indicates the penalty value for not meeting the maximum hanging quantity constraint of the workstation;

[0042] The objective function of minimizing the total load of the workstation f 2, its expression is:

[0043] ,

[0044] In the above formula, t ij For orders O i No. j The standard processing time of each process, p ijk For workstation k Processing Orders O i No. j The proficiency coefficient of the process, x ijk It is the station processing identification parameter, which is defined as follows:

[0045]

[0046] The objective function of minimizing the total transportation distance is f 3, whose expression is:

[0047] ,

[0048] In the above formula, d Indicates the transport distance between two workstations. W ij and W i(j+1) Processing Orders O i No. j The first process and the j+1 The workstation assigned to a process.

[0049] In some embodiments, in step S3, a penalty mechanism is added to increase the penalty value for the objective function value of the solution that does not meet the maximum hanging quantity constraint of the workstation, and the expression is:

[0050] ,

[0051] In the above formula, λ is the penalty coefficient, which ensures that individuals that violate the maximum hanging amount are eliminated during the optimization process; Indicates processing order i Time workstation k Whether the maximum hanging quantity constraint of the workstation is violated.

[0052] In some embodiments, the method for determining whether the maximum hanging quantity constraint of a workstation is violated is:

[0053] Calculate the material arrival rate at the workstation, which represents the processing order O i Time per unit time from workstation k -1 Arrival at workstation k The material quantity of the workstation is calculated as follows:

[0054] ,

[0055] In the above formula, For workstation k -1Processing order O i The process completion time is For processing orders O i Time workstation k -1 to the workstation k The transportation time of the workstation is calculated by the formula Perform calculations, v is the hanger moving speed;

[0056] Calculate the material processing rate of the workstation, which represents the processing order O i Workstation per unit time k The quantity of materials that can be completed is calculated as follows:

[0057] ;

[0058] Calculate the order completion time, which estimates the order completion time by calculating the processing time and transportation time. The calculation formula is as follows:

[0059] ,

[0060] According to the scheduling plan, the processing time of each workstation corresponding to each order process is calculated. t 0 is the initial processing time, Indicates order i The processing time increases linearly, with the maximum being the time to complete the entire order; here Indicates the processing steps j To processing steps j +1 transportation time between workstations; For orders O i The number of processing increases linearly, with the maximum being the total number of processing for this order Q i ;

[0061] Calculate the material accumulation at the workstation, which represents the material accumulation at each moment during the production of each order. The calculation formula is as follows:

[0062] ;

[0063] Determine the workstation k Whether the material accumulation during the production process will be greater than the maximum hanging capacity of the workstation, the judgment method is as follows:

[0064]

[0065] The above judgment method indicates that in the order O i During the production process, as the number of q Increase in workstations k Material accumulation It keeps increasing. Once it is greater than the maximum hanging amount of the workstation, it is judged as 1, which means that it does not meet the maximum hanging amount constraint of the material.

[0066] In some embodiments, in step S3, the constraints include:

[0067] The unique station allocation constraint is used to ensure that each process is completed at only one station. The expression is:

[0068] ,

[0069] In the above formula, x ijk Indicates order i No. j Is the first process at the workstation? k Processing, 1 if processing, otherwise 0, K is the total number of workstations;

[0070] The process-station matching constraint is used to ensure that each process can only be assigned to a station that meets the processing requirements. The expression is:

[0071] ;

[0072] The maximum hanging amount constraint of the workstation is used to ensure that the hanging amount of the workstation during the production process is not greater than the maximum hanging amount set by itself. The expression is:

[0073] ,

[0074] In the above formula, h ik Indicates workstation k Processing Orders i The amount of hanging at the time.

[0075] In some embodiments, in step S4, the improved non-dominated sorting genetic algorithm includes:

[0076] Step S41, setting population optimization parameters, using a three-layer coding method of process and station with detour marks to encode population individuals, the first layer is the process code, the second layer is the station code, and the third layer is the detour mark, which indicates the number of detours required to reach the station of the next process after the station completes the corresponding process;

[0077] Step S42, when generating the initial population, three strategies including path optimization, heuristic based on maximum completion time and random selection are adopted;

[0078] Step S43, calculating the objective function value of the individuals in the population, and performing non-dominated sorting and crowding distance calculation on the individuals;

[0079] Step S44, based on the non-dominated sorting and crowding distance, a certain number of optimal individuals are selected as elite individuals for retention;

[0080] Step S45, grouping the population according to the optimization target in the optimization stage, and performing crossover operation between groups on the individuals of the population, wherein the crossover operation adopts a direction-sensitive crossover method;

[0081] Step S46, calculate the similarity of the individuals in the population, construct an individual similarity matrix, use a voting mechanism to determine whether the population is trapped in a local optimum, and trigger individual mutation when the mutation condition is met, and mutate the highly similar individuals, so that the population jumps out of the local optimum and expands the search to the global optimum;

[0082] Step S47, merging the elite individuals of the parent generation and the offspring individuals after crossover mutation to form a new population;

[0083] Step S48, repeat algorithm steps S43-S47 until the number of optimization iterations reaches the maximum number of iterations, and output the Pareto frontier solution set of the population.

[0084] In some embodiments, in step S42, initialization of path optimization includes:

[0085] Step S421, define the state space, the expression is:

[0086]

[0087] In the above formula, current Indicates the current workstation position.next Indicates the location of the next process. path_history Indicates the workstation path that has been passed. work_history Indicates the number of visits to each station on the processing path, total_distance Indicates the current accumulated transportation distance;

[0088] Step S422, define the process optional station array, according to the order process and process station matching matrix C , establish each process of the order next An array whose expression is:

[0089]

[0090] In the above formula, next i Indicates i At which stations can the process be processed?

[0091] Step S423, define a heuristic function, whose expression is:

[0092] ,

[0093] In the above formula, D is the workstation distance matrix, current For the current workstation, goal For the target workstation, μ represents the penalty weight of the number of detours, count Indicates the number of detours;

[0094] Step S424, using path optimization to initialize the population, including path initialization, station expansion, state space update, and termination conditions.

[0095] Path initialization, set a priority queue of size F to store the path, add the initial state space to the priority queue, and its expression is:

[0096] ( start, next 2, [], [0...0], 0),

[0097] In the above formula, start For next The station number selected in 1 is the station for processing the first process. next 2 is the starting station after processing, the second process can be processed in which stations, the initial path_history is empty, work_history The number of visits to all workstations is 0. total_distance is 0.

[0098] Expand the workstation and expand the state space in the priority queue. Each state spacenext Array to select station expansion, select work_history All workstations that have been visited no more than once.

[0099] Update the state space. For each expanded state space, update current For next The selected workstation number will next The array is updated to the next process next Array, add the selected station number to path_history In work_history The number of visits to the corresponding workstation number in is increased by 1, and the calculation is total_distance = total_distance + h (x) . The updated state space will be expanded according to total_distance Sort by size, select the first F smallest state spaces and add them to the priority queue.

[0100] Termination conditions, NEXT All arrays in the priority queue are visited in order, that is, all processes select a station for processing, and the priority queue of each state space is path_history Used for station initialization of scheduling plan.

[0101] In some embodiments, in step S42, the heuristic initialization based on the maximum completion time is specifically:

[0102] For processes that can be processed at multiple stations, the strategy for allocating stations is as follows: First, when the processing order has not started, the load of all stations is 0. At this time, according to the machine type required to process the process, select the machine type from the station machine type array. W Get all the station numbers of the corresponding machine types, and calculate them according to the worker proficiency coefficient matrix P， Select the workstation with the highest proficiency coefficient for the process and assign it to the workstation. After that, each process that uses the machine type will select an unselected workstation according to the proficiency coefficient until all workstations of the machine type have been selected.

[0103] After all the stations of the same machine type are selected, if there is a subsequent process that needs to use this type of machine for processing, the load and proficiency evaluation function of all stations in this machine type is calculated first. The calculation formula is as follows:

[0104]

[0105] in, t i,k Indicates workstation k Processing i The time of the process is used to express the load. p i,k Indicates at workstation kProcessing i The proficiency coefficient corresponding to the process, n k Indicates at workstation k The number of processing steps. In this expression, if the station k Load t i,k The larger the value, the higher the evaluation score. E k The smaller the proficiency coefficient p i,k The higher the score, the E k The larger the number, the more statistical stations k All the processes in process are calculated to get the workstation k Load and proficiency assessment scores;

[0106] All workstations of the same machine type are divided into E k Sort, each time you select a station based on machine type, choose the largest E k The corresponding workstations are those with lighter loads and higher proficiency coefficients.

[0107] In some embodiments, in step S45, the algorithm optimization stage includes an early optimization stage and a late optimization stage; in the early optimization stage, the three optimization goals of completion time, workstation load, and total transportation distance are sorted according to the optimization objective function value, and the population individuals are grouped into three intervals of high, medium, and low; in the late optimization stage, the three optimization goals of completion time, total transportation distance, and workstation load are respectively selected, and popSize / 3 optimal individuals are grouped.

[0108] In some embodiments, in step S46, the similarity calculation formula is:

[0109] ;

[0110] in, , n is the total length of the code, ns The number of characters that are exactly the same at the same position in the process level code; , count Assign the same number of workstations to the workstations under the same order process in the workstation layer coding.

[0111] In some embodiments, the similarity matrix is:

[0112]

[0113] In the above formula, M [ i ][j ] = S [ i ][ j ], M [ i ][ j ]=1 means that the similarity between an individual and itself is 1. M [ i ][ j ]The closer it is to 1, the more similar the two individuals are, and the closer it is to 0, the more different the two individuals are.

[0114] In some embodiments, the voting mechanism is specifically:

[0115] Setting similarity threshold , if the individual i and individuals j The similarity value of S ( i , j ) exceeds the set threshold , then the two individuals are considered similar;

[0116] According to the similarity matrix, each individual is voted. The voting method is to calculate the number of similar individuals between each individual and other individuals. The calculation formula is as follows:

[0117]

[0118] in, V i represents the voting score of an individual, is a representation function, when 1 when it is, otherwise 0;

[0119] If the individual i Voting score V i If more than 0.5 popSize is exceeded, that is, more than half of the individuals are similar to it, then the individual is considered i It is a highly similar individual and is part of the local optimal population; the number of highly similar individuals in the entire population is counted. When the number of highly similar individuals accounts for 60%-80% of the population, it is considered that the population has fallen into the local optimum and a mutation operation is required. When the proportion is less than 60% or greater than 80% and the average improvement in fitness value is less than 1%, no mutation operation is required;

[0120] The fitness improvement calculation formula is as follows:

[0121]

[0122] in, F n (i ) represents the nth iteration i The fitness value of an individual.

[0123] The present invention also provides a clothing hanging production scheduling system taking into account the transportation distance, which is characterized by comprising:

[0124] An acquisition module, which is used to acquire information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production logs, and clothing hanging workshop layout;

[0125] A processing module, which is used to process the information of the clothing hanging production workshop to obtain the worker proficiency coefficient matrix and the workstation distance matrix of the hanging workshop;

[0126] A construction module is used to establish a mathematical model for clothing hanging production scheduling taking into account transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints;

[0127] The solution module is used to introduce an improved non-dominated sorting genetic algorithm to solve the mathematical model of clothing hanging production scheduling to obtain a solution for clothing hanging production scheduling.

[0128] By adopting the above technical solution, the beneficial effects of the present invention are:

[0129] The multi-objective optimization algorithm adopted by the present invention, in combination with the actual layout and production situation of the clothing hanging workshop, takes full account of the optimization of the running route of the hanging line hanger, introduces initialization strategies such as path optimization, and improves the cross-mutation operation, with a small number of optimization iterations, which can effectively avoid falling into the local optimal solution, has a strong multi-objective optimization capability, and has a high quality of solution. The present invention can well replace manual labor for clothing hanging production scheduling, optimize the generated scheduling plan, not only can reduce the production cycle, balance the load of the production line station, improve the workshop production efficiency and product quality, but also can effectively reduce costs, achieve more flexible and efficient production management, and enhance the market competitiveness of enterprises.

[0130] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure.

[0131] Undoubtedly, these and other objects of the present invention will become more apparent after the following detailed description of the preferred embodiment described with reference to various figures and drawings.

[0132] In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, one or several preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0133] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation of the present invention.

[0134] In the drawings, the same reference numerals are used for the same components and the drawings are schematic and not necessarily drawn to scale.

[0135] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only one or several embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on such drawings without paying creative work.

[0136] Figure 1 It is the scheduling flow chart of the present invention;

[0137] Figure 2 It is the algorithm flow chart of the present invention;

[0138] Figure 3 This is a schematic diagram of clothing hanging production and transportation;

[0139] Figure 4 Schematic diagram of coding for individuals in the population with detour marks;

[0140] Figure 5 To optimize the crossover diagram of the early population grouping;

[0141] Figure 6 To optimize the crossover diagram of population grouping in the later stage;

[0142] Figure 7 It is a schematic diagram of the partial mapping crossover method based on detour feedback;

[0143] Figure 8 This is the Gantt chart for scheduling scheme 1;

[0144] Figure 9 This is the Gantt chart for scheduling plan 2. DETAILED DESCRIPTION

[0145] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but are not used to limit the present invention.

[0146] In the present invention, unless otherwise clearly specified and limited, the first feature "on" or "under" the second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in an appropriate manner in any one or more embodiments or examples.

[0147] Reference Figure 1 , Figure 1 It is a scheduling flow chart of the present invention.

[0148] According to some embodiments of the present invention, the present invention provides a method and system for scheduling clothing hanging production taking transportation distance into consideration.

[0149] The technical solution adopted by the present invention to solve the technical problem is: a clothing hanging production scheduling method considering the transportation distance, comprising the following steps:

[0150] Step S1, obtaining information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production log, and clothing hanging workshop layout;

[0151] Step S2, processing the information of the garment hanging production workshop to obtain the worker proficiency coefficient matrix of the hanging workshop (calculating the worker proficiency coefficients assigned to different workstations in the processing procedures according to the hanger flow history data in the obtained information of the garment hanging production workshop) and the workstation distance matrix (establishing the workstation distance matrix according to the workstation distance in the obtained information of the garment hanging production workshop);

[0152] Step S3, establishing a mathematical model for clothing hanging production scheduling considering transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints;

[0153] Step S4, introducing an improved non-dominated sorting genetic algorithm to solve the established mathematical model of clothing hanging production scheduling, so as to obtain a solution for clothing hanging production scheduling.

[0154] According to some embodiments of the present invention, optionally, in step S1, obtaining information of a clothing hanging production workshop includes:

[0155] Collect and pre-process order information, process information and workstation information based on garment production information;

[0156] Collect the historical data of clothing hanging hanger flow according to the clothing hanging production log (the time it takes for each hanger to enter a certain workstation from the first processing step to the final completion and removal from the shelf t 进入 and departure time t 离开 , and the number of repairs entering this station N r );

[0157] Collect the distance between workstations according to the layout of the clothing hanging workshop d and hanger moving speed v .

[0158] According to some embodiments of the present invention, optionally, pre-processing order information, process information and workstation information is specifically as follows:

[0159] The processing time matrix is ​​established based on the standard processing time of the order process in the obtained process information. Its expression is:

[0160]

[0161] In the above formula, n is the number of orders, m for n The maximum number of processes in an order;

[0162] Create a workstation machine type array based on the workstation number and machine type in the workstation information. The expression is:

[0163]

[0164] In the above formula, k is the largest workstation number, t is the largest machine type number, w ij The value is 0 or 1. w ij =1 indicates workstation i The machine type is j , a workstation can only use machines of one machine type, i.e. ;

[0165] According to the maximum hanging amount in the obtained workstation information, the workstation maximum hanging amount array is established, and its expression is:

[0166]

[0167] According to the processing matching of the obtained order process and workstation, a process-workstation matching matrix is ​​established, and its expression is:

[0168]

[0169] In the above formula, m i For orders O i The number of processes, c ijk Indicates order O i No. j Is the process at the workstation? k If it can be processed, it is 1, otherwise it is 0.

[0170] In one embodiment of the present invention, it is assumed that the order processing details table of a clothing hanging production workshop is as shown in Table 1:

[0171] Table 1 Order processing details

[0172]

[0173] Then in this embodiment, the processing time matrix is:

[0174]

[0175] The detailed list of workstation resources in the clothing hanging production workshop is shown in Table 2:

[0176] Table 2 Workstation resource details

[0177]

[0178] Then the workstation machine type array in this embodiment is:

[0179]

[0180] The machine types for lockstitch sewing machine, buttonhole machine and ironing machine are 1, 2 and 3 respectively;

[0181] The maximum hanging amount array is:

[0182] ,

[0183] The process station matching matrix is:

[0184] .

[0185] According to some embodiments of the present invention, optionally, in step S2, processing the information of the clothing hanging production workshop includes:

[0186] According to the hanger flow history data obtained from the clothing hanging production workshop information, the proficiency coefficients of the designated workers at different workstations in the processing procedures are calculated, and the worker proficiency coefficient matrix is ​​statistically established. The expression is:

[0187]

[0188] In the above formula, p ijk Indicates workstation k Processing Orders O i No. j The proficiency coefficient of a process is expressed as:

[0189] ,

[0190] In the above formula, α is the efficiency weight factor, β is the quality weight factor, satisfying α + β =1, the enterprise can determine it according to production strategy and market demand, and make corresponding adjustments according to the efficiency and quality requirements of different customers and orders; t ijk Indicates workstation k Processing Orders O i No. j The actual processing time of a process is calculated by collecting historical data on the flow of hangers to calculate the time when hangers enter and leave the workstation during the process. k time difference; t ij Indicates processing order O i No. j The standard processing time of each process; N i For orders O i The total production quantity of N r For workstation k The amount of rework required to process this operation; t ijk The calculation formula is:

[0191] ,

[0192] In the above formula, t 进入 , t 离开 Workstations k Processing Orders O i No. j Each hanger enters and leaves the workstation during the process k time.

[0193] In one embodiment of the present invention, it is assumed that the proficiency coefficient table obtained by collection and calculation is as shown in Table 3.

[0194] Table 3 Proficiency coefficient table

[0195]

[0196] Then the proficiency coefficient matrix is:

[0197] .

[0198] According to some embodiments of the present invention, optionally, the clothing hanging layout currently generally uses a U-shaped layout such as Figure 3 As shown in the figure, for the stations on the same side, the distance between the stations is the same, and the distance is defined directly according to the sequence and the track length. For crossing the U-shaped bend to enter the station on the opposite side, the track length of the bend needs to be defined. The transportation of the U-shaped hanging production line can only run in one direction, so the direction restriction needs to be considered when calculating the path, and the distance matrix is ​​established using the adjacency matrix method. The station distance matrix is ​​expressed as:

[0199] ,

[0200] In the above formula, d is the distance between adjacent workstations on the same side, and is the distance between the adjacent stations of the U-shaped bend of clothing hanging. The above matrix assumes that there are 4 stations in the hanging workshop, stations 1, 2, 3, and 4 are on the same side, so the distance from station 1 to station 3 is , and because the transportation direction is unidirectional, station 2 needs to go around to stations 3 and 4 to reach station 1, and the distance is .

[0201] In one embodiment of the present invention, it is assumed that the distance between each station is:

[0202] , the distance in m, and the speed of the hanger v =0.5m / s.

[0203] According to some embodiments of the present invention, optionally, in step S3, optimizing the objective function includes:

[0204] The objective function to minimize the maximum completion time is f 1, whose expression is:

[0205] ,

[0206] In the above formula, C max Indicates the maximum completion time for processing all orders. Ci Indicates order i The completion time, n Indicates the total number of orders, R Indicates the penalty value for not meeting the maximum hanging quantity constraint of the workstation;

[0207] The objective function of minimizing the total load of the workstation f 2, its expression is:

[0208] ,

[0209] In the above formula, t ij For orders O i No. j The standard processing time of each process, p ijk For workstation k Processing Orders O i No. j The proficiency coefficient of the process, x ijk It is the station processing identification parameter, which is defined as follows:

[0210]

[0211] The objective function of minimizing the total transportation distance is f 3, whose expression is:

[0212] ,

[0213] In the above formula, d Indicates the transport distance between two workstations. W ij and W i(j+1) Processing Orders O i No. j The first process and the j+1 The workstation assigned to a process.

[0214] According to some embodiments of the present invention, optionally, in step S3, a penalty mechanism is added to increase a penalty value for the objective function value of the solution that does not meet the constraint of the maximum hanging amount of the workstation, and the expression is:

[0215] ,

[0216] In the above formula, λ is the penalty coefficient, which ensures that individuals that violate the maximum hanging amount are eliminated during the optimization process; Indicates processing order i Time workstation kWhether the maximum hanging quantity constraint of the workstation is violated.

[0217] Since the hanging amount of each workstation cannot be greater than the set maximum hanging amount, otherwise it will cause materials to accumulate on the main hanging track, causing the operation of the main track to be blocked. Therefore, a penalty mechanism is adopted to deal with the material accumulation situation. A higher penalty is imposed on the individual objective function value that does not meet the maximum hanging amount constraint of the workstation, so that it is eliminated in the optimization process, thereby ensuring that the final solution meets the maximum hanging amount constraint and speeding up the acquisition of the optimal solution.

[0218] According to some embodiments of the present invention, optionally, a method for determining whether a maximum hanging quantity constraint of a workstation is violated is:

[0219] Calculate the material arrival rate at the workstation, which represents the processing order O i Time per unit time from workstation k -1 Arrival at workstation k The material quantity of the workstation is calculated as follows:

[0220] ,

[0221] In the above formula, For workstation k -1Processing order O i The process completion time is For processing orders O i Time workstation k -1 to the workstation k The transportation time of the workstation is calculated by the formula Perform calculations, v is the hanger moving speed;

[0222] Calculate the material processing rate of the workstation, which represents the processing order O i Workstation per unit time k The quantity of materials that can be completed is calculated as follows:

[0223] ;

[0224] Calculate the order completion time, which estimates the order completion time by calculating the processing time and transportation time. The calculation formula is as follows:

[0225] ,

[0226] According to the scheduling plan, the processing time of each workstation corresponding to each order process is calculated. t 0 is the initial processing time, Indicates order iThe processing time increases linearly, with the maximum being the time to complete the entire order; here Indicates the processing steps j To processing steps j +1 transportation time between workstations; For orders O i The number of processing increases linearly, with the maximum being the total number of processing for this order Q i ;

[0227] Calculate the material accumulation at the workstation, which represents the material accumulation at each moment during the production of each order. The calculation formula is as follows:

[0228] ;

[0229] Determine the workstation k Whether the material accumulation during the production process will be greater than the maximum hanging capacity of the workstation, the judgment method is as follows:

[0230]

[0231] The above judgment method indicates that in the order O i During the production process, as the number of q Increase in workstations k Material accumulation It keeps increasing. Once it is greater than the maximum hanging amount of the workstation, it is judged as 1, which means that it does not meet the maximum hanging amount constraint of the material.

[0232] According to some embodiments of the present invention, optionally, in step S3, the constraint condition includes:

[0233] The unique station allocation constraint is used to ensure that each process is completed at only one station. The expression is:

[0234] ,

[0235] In the above formula, x ijk Indicates order i No. j Is the first process at the workstation? k Processing, 1 if processing, otherwise 0, K is the total number of workstations;

[0236] The process-station matching constraint is used to ensure that each process can only be assigned to a station that meets the processing requirements. The expression is:

[0237] ;

[0238] The maximum hanging amount constraint of the workstation is used to ensure that the hanging amount of the workstation during the production process is not greater than the maximum hanging amount set by itself. The expression is:

[0239] ,

[0240] In the above formula, h ik Indicates workstation k Processing Orders i The amount of hanging at the time.

[0241] Reference Figure 2 , Figure 2 It is the algorithm flow chart of the present invention.

[0242] According to some embodiments of the present invention, optionally, in step S4, the improved non-dominated sorting genetic algorithm includes:

[0243] Step S41, setting population optimization parameters, using a three-layer coding method with detour marks to encode population individuals, the first layer is the process code, the second layer is the station code, and the third layer is the detour mark, which indicates the number of times the station needs to detour to the next process after the corresponding process is completed.

[0244] The population optimization parameters mainly include the population size popSize and the maximum number of iterations of population optimization maxGen.

[0245] In one embodiment of the present invention, the population size popSize=50 is initialized, and the maximum number of iterations maxGen=100.

[0246] Since the one-way production line of the U-shaped layout of clothing hanging has the characteristic of detour, the material may need to complete one or more circles before returning to the starting point or a specific workstation during the hanging production process. Therefore, the coding method of detour avoidance is used, such as Figure 4 As shown, a layer of detour mark is added on the basis of the double-layer coding of the process station, and a detour mark is added to each process to indicate whether the process needs to detour to reach the processing station of the next process.

[0247] Figure 4 In the coding example shown, the first layer is the process code, which determines the processing order of the process, and the digital codes 1, 2, and 3 represent the order number; the second layer is the station code, which determines the processing station corresponding to each process, and the digital codes 1-5 represent the station number; the third layer is the detour mark, which indicates that a detour is required from the processing of this process to the processing of the next process, and the size of the number indicates the number of detours. Taking the number 2 in the first layer as an example, the first appearance of the number 2 represents the first process of order 2 O 21The second level number 2 at the corresponding position indicates that the processing is done at station 2, and the third level number 1 at the corresponding position indicates that the processing at station 2 is completed and the next process is reached. O 22 The corresponding processing station 1 needs to detour once.

[0248] The third layer of encoding in this encoding method serves as a feedback mechanism and does not participate in the subsequent crossover mutation process. It mainly functions to identify unreasonable workstation allocation and guide subsequent repair and local search operations to ensure that the newly generated individuals will not consume too much transportation distance due to unreasonable workstation allocation.

[0249] Step S42, when generating the initial population, three strategies are adopted, including path optimization, heuristic based on maximum completion time, and random selection.

[0250] For the optimization of transportation distance and maximum completion time, path optimization, heuristic based on maximum completion time and random selection are used to initialize the population respectively to improve the quality of the initial solution.

[0251] Step S43, calculating the objective function value of the individuals in the population, and performing non-dominated sorting and crowding distance calculation on the individuals.

[0252] Calculate the objective function values ​​of all individuals in the population, perform non-dominated sorting on all individuals in the current population, divide the population into multiple non-dominated layers, and calculate the crowding distance of individuals in each non-dominated layer.

[0253] Step S44, based on the non-dominated sorting and crowding distance, a certain number of optimal individuals are selected as elite individuals for retention.

[0254] According to the above non-dominated sorting and crowding distance, individuals with low non-dominated layers are given priority. In the same non-dominated layer, individuals with high crowding are given priority. A certain number of optimal individuals are selected as elite individuals and retained to build the next generation of new population.

[0255] Step S45, grouping the population according to the optimization target in the optimization stage, and performing crossover operation between groups on the individuals of the population, wherein the crossover operation adopts a direction-sensitive crossover method.

[0256] Step S46, calculate the similarity of the individuals in the population, construct an individual similarity matrix, and use a voting mechanism to determine whether the population is trapped in a local optimum. If the mutation conditions are met, individual mutation is triggered, and highly similar individuals are mutated, so that the population jumps out of the local optimum and expands the search to the global optimum.

[0257] During the optimization process, by calculating the similarity of individuals and using the voting mechanism, it is possible to effectively determine whether the population has fallen into a local optimum, and take targeted mutation measures to avoid premature convergence.

[0258] Step S47, merging the elite individuals of the parent generation and the offspring individuals after crossover mutation to form a new population.

[0259] The parent elite individuals selected above and the offspring individuals after the crossover operation and the mutation operation above form a new population.

[0260] Step S48, repeat algorithm steps S43-S47 until the number of optimization iterations reaches the maximum number of iterations, and output the Pareto frontier solution set of the population.

[0261] According to some embodiments of the present invention, optionally, in step S42, initialization of path optimization includes:

[0262] Step S421, define the state space, the expression is:

[0263]

[0264] In the above formula, current Indicates the current workstation position. next Indicates the location of the next process. path_history Indicates the workstation path that has been passed. work_history Indicates the number of visits to each station on the processing path, total_distance Indicates the current accumulated transportation distance;

[0265] Step S422, define the process optional station array, according to the order process and process station matching matrix C , establish each process of the order next An array whose expression is:

[0266]

[0267] In the above formula, next i Indicates i At which stations can the process be processed?

[0268] Step S423, define a heuristic function, whose expression is:

[0269] ,

[0270] In the above formula, D is the workstation distance matrix, current For the current workstation, goal For the target workstation, μ represents the penalty weight of the number of detours, count Indicates the number of detours;

[0271] Step S424, using path optimization to initialize the population, including path initialization, station expansion, state space update, and termination conditions.

[0272] Path initialization, set a priority queue of size F to store the path, add the initial state space to the priority queue, and its expression is:

[0273] ( start, next 2, [], [0...0], 0),

[0274] In the above formula, start For next The station number selected in 1 is the station for processing the first process. next 2 is the starting station after processing, the second process can be processed in which stations, the initial path_history is empty, work_history The number of visits to all workstations is 0. total_distance is 0.

[0275] Expand the workstation and expand the state space in the priority queue. Each state space next Array to select station expansion, select work_history All workstations that have been visited no more than once.

[0276] Update the state space. For each expanded state space, update current For next The selected workstation number will next The array is updated to the next process next Array, add the selected station number to path_history In work_history The number of visits to the corresponding workstation number in is increased by 1, and the calculation is total_distance = total_distance + h (x) . The updated state space will be expanded according to total_distance Sort by size, select the first F smallest state spaces and add them to the priority queue.

[0277] Termination conditions, NEXT All arrays in the priority queue are visited in order, that is, all processes select a station for processing, and the priority queue of each state space is path_history Used for station initialization of scheduling plan.

[0278] According to some embodiments of the present invention, optionally, in step S42, the heuristic initialization based on the maximum completion time is specifically:

[0279] For processes that can be processed at multiple stations, the strategy for allocating stations is as follows: First, when the processing order has not started, the load of all stations is 0. At this time, according to the machine type required to process the process, select the machine type from the station machine type array.W Get all the station numbers of the corresponding machine types, and calculate them according to the worker proficiency coefficient matrix P， Select the workstation with the highest proficiency coefficient for the process and assign it to the workstation. After that, each process that uses the machine type will select an unselected workstation according to the proficiency coefficient until all workstations of the machine type have been selected.

[0280] After all the stations of the same machine type are selected, if there is a subsequent process that needs to use this type of machine for processing, the load and proficiency evaluation function of all stations in this machine type is calculated first. The calculation formula is as follows:

[0281]

[0282] in, t i,k Indicates workstation k Processing i The time of the process is used to express the load. p i,k Indicates at workstation k Processing i The proficiency coefficient corresponding to the process, n k Indicates at workstation k The number of processing steps. In this expression, if the station k Load t i,k The larger the value, the higher the evaluation score. E k The smaller the proficiency coefficient p i,k The higher the score, the E k The larger the number, the more statistical stations k All the processes in process are calculated to get the workstation k Load and proficiency assessment scores;

[0283] All workstations of the same machine type are divided into E k Sort, each time you select a station based on machine type, choose the largest E k The corresponding workstations are those with lighter loads and higher proficiency coefficients.

[0284] According to some embodiments of the present invention, optionally, the optimization stage includes an early optimization stage and a late optimization stage. Two grouping strategies are used in the two periods before and after the optimization stage to group the population according to the optimization target, and the maximum number of iterations maxGen is set, with the first 60% as the early optimization stage and the last 40% as the late optimization stage.

[0285] In the early stage of optimization, the three optimization goals of completion time, total transportation distance, and station load are sorted according to the optimization objective function values, and the individuals in the population are grouped into three intervals: high, medium, and low. For example, for the optimization goal of completion time, the individuals are sorted according to the objective function value, and are divided into three groups: short time, medium time, and long time. The crossover strategy after grouping is to crossover the high and low objective function values ​​between different optimization goals, and crossover the middle part of the objective function values. Figure 5 This approach can better balance local development and global exploration, ensuring that the algorithm can develop towards the optimal solution while retaining the diversity of the population.

[0286] In the later stage of optimization, popSize / 3 optimal individuals are selected according to the three optimization objectives of completion time, total transportation distance, and workstation load. The crossover strategy after grouping is to perform pairwise crossover between groups with different optimization objectives, such as Figure 6 In the later stages of optimization, the global exploration requirements of the problem are reduced, and more attention is paid to local fine-tuning. Therefore, in this way, the selected individuals are all high-quality solutions close to the optimal solution. Pairwise crossover can more effectively help the solution approach the global optimal solution, thereby accelerating the convergence speed.

[0287] The crossover method uses a partial mapping crossover method based on detour feedback, and refers to the third-level detour mark of the parent before crossing, such as Figure 7 As shown, the specific crossover steps are as follows:

[0288] 1. First, code and split the parent generation to be crossed, and divide the process and workstation codes of the parent generation into three parts: head, middle and tail.

[0289] 2. For each segment of the two parent generations, count the third-level detour marks, calculate the number of detours for all workstations in the segment, compare the number of detours in the three segments of the parent generations, and select the segment with the largest difference.

[0290] 3. Cross the section with the largest difference in the number of detours, and retain the other two coded sections with fewer detours. The crossed sections may lead to repeated processes, which are processed using the conventional partial mapping crossover method, that is, the repeated processes are replaced with the missing processes in turn. If the workstation is incompatible after the process is replaced, select one from the workstations that are compatible with the process and do not cause detours.

[0291] 4. After executing the crossover, recalculate the detour mark of the third layer and update the new individuals generated by the crossover.

[0292] According to some embodiments of the present invention, optionally, in step S46, the similarity calculation formula is:

[0293] ;

[0294] in, , n is the total length of the code, ns The number of characters that are exactly the same at the same position in the process level code; , count Assign the same number of workstations to the workstations under the same order process in the workstation layer coding.

[0295] According to some embodiments of the present invention, optionally, the similarity matrix is:

[0296]

[0297] In the above formula, M [ i ][ j ] = S [ i ][ j ], M [ i ][ j ]=1 means that the similarity between an individual and itself is 1. M [ i ][ j ]The closer it is to 1, the more similar the two individuals are, and the closer it is to 0, the more different the two individuals are.

[0298] According to some embodiments of the present invention, optionally, the voting mechanism is specifically:

[0299] Setting similarity threshold , if the individual i and individuals j The similarity value of S ( i , j ) exceeds the set threshold , then the two individuals are considered similar;

[0300] According to the similarity matrix, each individual is voted. The voting method is to calculate the number of similar individuals between each individual and other individuals. The calculation formula is as follows:

[0301]

[0302] in, V i represents the voting score of an individual, is a representation function, when 1 when it is, otherwise 0;

[0303] If the individual i Voting score V iIf more than 0.5 popSize is exceeded, that is, more than half of the individuals are similar to it, then the individual is considered i It is a highly similar individual and is part of the local optimal population; the number of highly similar individuals in the entire population is counted. When the number of highly similar individuals accounts for 60%-80% of the population, it is considered that the population has fallen into the local optimum and a mutation operation is required. When the proportion is less than 60% or greater than 80% and the average improvement in fitness value is less than 1%, no mutation operation is required;

[0304] The fitness improvement calculation formula is as follows:

[0305]

[0306] in, F n ( i ) represents the nth iteration i The fitness value of an individual.

[0307] In one embodiment of the present invention, the Pareto frontier solution set obtained above is decoded to generate a clothing hanging production scheduling plan, and compared with the plan without considering the transportation distance, as shown in Table 4, including the processing sequence of the process and the processing station corresponding to the process, where scheduling plan 1 considers the transportation distance, and scheduling plan 2 does not consider the transportation distance. The corresponding scheduling plan comparison Gantt chart is shown in Figure 4. Figures 8 - 9 The objective function values ​​of the two scheduling schemes are compared as shown in Table 5. From the data in the table, it can be found that the scheduling scheme that does not consider the transportation distance has a longer completion time in the actual production process due to the longer transportation time. The longer total transportation distance also increases the production energy consumption to a certain extent.

[0308] Table 4 Comparison of clothing hanging production scheduling schemes

[0309]

[0310] Table 5. Objective function values ​​of scheduling comparison schemes

[0311]

[0312] The present invention also provides a clothing hanging production scheduling system taking into account the transportation distance, which is characterized by comprising:

[0313] An acquisition module, which is used to acquire information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production logs, and clothing hanging workshop layout;

[0314] A processing module, which is used to process the information of the clothing hanging production workshop to obtain the worker proficiency coefficient matrix and the workstation distance matrix of the hanging workshop;

[0315] A construction module is used to establish a mathematical model for clothing hanging production scheduling taking into account transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints;

[0316] The solution module is used to introduce an improved non-dominated sorting genetic algorithm to solve the mathematical model of clothing hanging production scheduling to obtain a solution for clothing hanging production scheduling.

[0317] It should be understood that the embodiments disclosed in the present invention are not limited to the specific processing steps or materials disclosed herein, but should be extended to equivalent substitutions of such features understood by ordinary technicians in the relevant field. It should also be understood that the terms used herein are only used for the purpose of describing specific embodiments and are not meant to be limiting.

[0318] The "embodiment" mentioned in the specification means that the specific features or characteristics described in conjunction with the embodiment are included in at least one embodiment of the present invention. Therefore, the phrases or "embodiment" appearing in various places throughout the specification do not necessarily refer to the same embodiment.

[0319] In addition, the described features or characteristics may be combined in one or more embodiments in any other suitable manner. In the above description, some specific details, such as thickness, quantity, etc., are provided to provide a comprehensive understanding of the embodiments of the present invention. However, those skilled in the relevant art will understand that the present invention can be implemented without one or more of the above specific details or can also be implemented using other methods, components, materials, etc.

Claims

1. A method for scheduling clothing hanging production taking into account transportation distance, characterized in that: The following steps are involved: Step S1, obtaining information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production log, and clothing hanging workshop layout; Step S2, processing the information of the garment hanging production workshop to obtain the worker proficiency coefficient matrix and the workstation distance matrix of the hanging workshop; Step S3, establishing a mathematical model for clothing hanging production scheduling considering transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints; Step S4, introducing an improved non-dominated sorting genetic algorithm to solve the established mathematical model of clothing hanging production scheduling to obtain a solution for clothing hanging production scheduling; In step S2, the information of the clothing hanging production workshop is processed, including: According to the hanger flow history data obtained from the clothing hanging production workshop information, the proficiency coefficients of the designated workers at different workstations in the processing procedures are calculated, and the worker proficiency coefficient matrix is ​​statistically established. The expression is: In the above formula, p ijk Indicates that workstation k processes order O i The proficiency coefficient of the jth process is expressed as: In the above formula, α is the efficiency weight factor, β is the quality weight factor, and if α+β=1, the enterprise can determine it according to production strategy and market demand, and make corresponding adjustments according to the efficiency and quality requirements of different customers and orders; t ijk Indicates that workstation k processes order O i The actual processing time of the jth process is calculated by collecting the historical data of hanger flow and counting the time difference between the hanger entering and leaving the workstation k during the process; t ij Indicates processing order O i The standard processing time of the jth process; N i For order O i The total production quantity, and N r is the number of reworks required when processing this process at station k; t ijk The calculation formula is: In the above formula, t 进入 ,t 离开 They are processing orders O at station k respectively i The time for each hanger to enter and leave station k during the j-th process; According to the distance between workstations in the obtained clothing hanging production workshop information, the workstation distance matrix is ​​established using the adjacency matrix method, and the expression is: In the above formula, d is the distance between adjacent workstations on the same side, and d* is the distance between adjacent workstations in the U-bend of clothing hanging. The above matrix assumes that there are 4 workstations in the hanging workshop, and workstations 1, 2, 3, and 4 are on the same side. Therefore, the distance from workstation 1 to workstation 3 is d+d*. Due to the unidirectional nature of transportation, workstation 2 to workstation 1 needs to detour to workstations 3 and 4 to reach the destination, and the distance is d+2d*.

2. The method for scheduling clothing hanging production considering transportation distance according to claim 1 is characterized in that: In step S1, the information of the clothing hanging production workshop is obtained, including: Collect and pre-process order information, process information and workstation information based on garment production information; Collect clothing hanging hanger flow history data based on clothing hanging production logs; The distance between workstations and the speed of hanger movement are collected according to the layout of the clothing hanging workshop.

3. The method for scheduling clothing hanging production considering transportation distance according to claim 2 is characterized in that: Pre-process order information, process information and workstation information, specifically: The processing time matrix is ​​established based on the standard processing time of the order process in the obtained process information. Its expression is: In the above formula, n is the number of orders, and m is the maximum number of processes in n orders; Create a workstation machine type array based on the workstation number and machine type in the workstation information. The expression is: In the above formula, k is the largest station number, t is the largest machine type number, and w is the largest machine type number. ij The value is 0 or 1, w ij =1 means that the machine type of workstation i is j, and a workstation can only use machines of one machine type, that is, According to the maximum hanging amount in the obtained workstation information, the workstation maximum hanging amount array is established, and its expression is: According to the processing matching of the obtained order process and workstation, a process-workstation matching matrix is ​​established, and its expression is: In the above formula, m i For order O i The number of processes, c ijk Indicates order O i Whether the j-th process can be processed at the workstation k, if it can be processed, it is 1, otherwise it is 0.

4. The method for scheduling clothing hanging production considering transportation distance according to claim 1, characterized in that: In step S3, the optimization objective function includes: The objective function f1 that minimizes the maximum completion time is expressed as: f1=min(C max +R)=min{maxC i +R|i=1,2,...,n}, In the above formula, C max represents the maximum completion time for processing all orders, C i represents the completion time of order i, n represents the total number of orders, and R represents the penalty value for not meeting the maximum hanging quantity constraint of the workstation; The objective function f2 that minimizes the total load of the workstation is expressed as: In the above formula, t ij For order O i The standard processing time of the jth process, p ijk Processing order O for station k i The proficiency coefficient of the jth process, x ijk It is the station processing identification parameter, which is defined as follows: The objective function f3 that minimizes the total transportation distance is expressed as: In the above formula, d represents the transportation distance between two stations, W ij and W i(j+1) They are processing order O i The workstations assigned to the j-th process and the j+1-th process.

5. The method for scheduling clothing hanging production considering transportation distance according to claim 1, characterized in that: In step S3, a penalty mechanism is added to increase the penalty value for the objective function value of the solution that does not meet the maximum hanging quantity constraint of the workstation. The expression is: In the above formula, λ is the penalty coefficient, which ensures that individuals violating the maximum hanging amount are eliminated during the optimization process; φ i,k Indicates whether workstation k violates the maximum hanging quantity constraint of the workstation when processing order i.

6. The method for scheduling clothing hanging production considering transportation distance according to claim 5 is characterized in that: The judgment method for violating the maximum hanging quantity constraint of the workstation is: Calculate the material arrival rate at the workstation, which represents the processing order O i The number of materials that arrive at station k from station k-1 per unit time is calculated as follows: In the above formula, t k-1 Processing order O for station k-1 i The process completion time is For processing order O i The transportation time from station k-1 to station k is calculated by the formula Calculate, v is the speed of the hanger moving; Calculate the material processing rate of the workstation, which represents the processing order O i The number of materials that can be completed at workstation k per unit time is calculated as follows: Calculate the order completion time, which estimates the order completion time by calculating the processing time and transportation time. The calculation formula is as follows: According to the scheduling plan, the processing time of each workstation corresponding to each order process is calculated, where t0 is the initial processing time, Δt i represents the linear increase in the processing time of order i, with a maximum of the time required to process the entire order; here represents the transportation time between the processing step j and the processing step j+1; Δq is the order O i The number of processing increases linearly, with the maximum being the total number of processing for this order, Q i ; Calculate the material accumulation at the workstation, which represents the material accumulation at each moment during the production of each order. The calculation formula is as follows: Dh i,k =(δ i,k -c i,k )×Δt i ; Determine whether the material accumulation amount of station k will be greater than the maximum hanging amount of the station during the production process. The judgment method is as follows: The above judgment method indicates that in order O i During the production process, as the number of processes q increases, the amount of material accumulated at station k Δh ik It keeps increasing. Once it is greater than the maximum hanging amount of the workstation, it is judged as 1, which means that it does not meet the maximum hanging amount constraint of the material.

7. The method for scheduling clothing hanging production considering transportation distance according to claim 1, characterized in that: In step S3, the constraints include: The unique station allocation constraint is used to ensure that each process is completed at only one station. The expression is: In the above formula, x ijk Indicates whether the jth process of order i is processed at workstation k. If it is processed, it is 1, otherwise it is 0. K is the total number of workstations. The process-station matching constraint is used to ensure that each process can only be assigned to a station that meets the processing requirements. The expression is: x ijk =1&&c ijk =1; The maximum hanging amount constraint of the workstation is used to ensure that the hanging amount of the workstation during the production process is not greater than the maximum hanging amount set by itself. The expression is: In the above formula, h ik It indicates the hanging quantity when station k processes order i.

8. The method for scheduling clothing hanging production considering transportation distance according to claim 1, characterized in that: In step S4, the improved non-dominated sorting genetic algorithm includes: Step S41, setting population optimization parameters, using a three-layer coding method of process and station with detour marks to encode population individuals, the first layer is the process code, the second layer is the station code, and the third layer is the detour mark, which indicates the number of detours required to reach the station of the next process after the station completes the corresponding process; Step S42, when generating the initial population, three strategies including path optimization, heuristic based on maximum completion time and random selection are adopted; Step S43, calculating the objective function value of the individuals in the population, and performing non-dominated sorting and crowding distance calculation on the individuals; Step S44, based on the non-dominated sorting and crowding distance, a certain number of optimal individuals are selected as elite individuals for retention; Step S45, grouping the population according to the optimization target in the optimization stage, and performing crossover operation between groups on the individuals of the population, wherein the crossover operation adopts a direction-sensitive crossover method; Step S46, calculate the similarity of the individuals in the population, construct an individual similarity matrix, use a voting mechanism to determine whether the population is trapped in a local optimum, and trigger individual mutation when the mutation condition is met, and mutate the highly similar individuals, so that the population jumps out of the local optimum and expands the search to the global optimum; Step S47, merging the elite individuals of the parent generation and the offspring individuals after crossover mutation to form a new population; Step S48, repeat algorithm steps S43-S47 until the number of optimization iterations reaches the maximum number of iterations, and output the Pareto frontier solution set of the population.

9. The method for scheduling clothing hanging production considering transportation distance according to claim 8, characterized in that: In step S42, the initialization of the path optimization includes: Step S421, define the state space, the expression is: state=(current, next, path_history, work_history, total_distance) In the above formula, current represents the current workstation position, and next represents the workstation position that can be accessed in the next process. path_history indicates the workstation path that has been passed, and work_history indicates the number of times each workstation has been visited on the processing path. total_distance indicates the current accumulated transportation distance; Step S422, define the optional workstation array of the process, and establish the next array of each process of the order according to the process of the order and the process-workstation matching matrix C, the expression is: NEXT={next1,next2,...,next n }, In the above formula, next i Indicates which stations the i-th process can be processed at; Step S423, define a heuristic function, whose expression is: h(x)=D[current][goal]+μ×count, In the above formula, D is the workstation distance matrix, current is the current workstation, goal is the target workstation, μ represents the detour penalty weight, and count represents the detour times; Step S424, using path optimization to initialize the population, including path initialization, station expansion, state space update, and termination conditions; Path initialization, set a priority queue of size F to store the path, add the initial state space to the priority queue, and its expression is: (start, next2, [], [0...0], 0), In the above formula, start is the station number selected from next1, that is, the station where the first process is processed, next2 is the station where the second process can be processed after the starting station is processed, the initial path_history is empty, the number of visits to all stations in work_history is 0, and total_distance is 0; Expand the workstations, expand the state space in the priority queue, each state space selects the workstation expansion according to its own next array, and selects all workstations that have been accessed no more than once in work_history; Update the state space. For each expanded state space, update current to the workstation number selected from next, update the next array to the next array of the next process, add the selected workstation number to path_history, add 1 to the number of visits to the corresponding workstation number in work_history, and calculate total_distance = total_distance + h(x); sort the expanded and updated state spaces by total_distance, select the first F smallest state spaces and add them to the priority queue; Termination condition: all arrays in NEXT have been visited in sequence, that is, all processes have selected a workstation for processing, and the path_history of each state space in the priority queue is used to initialize the workstation of the scheduling plan.

10. The method for scheduling clothing hanging production considering transportation distance according to claim 8, characterized in that: In step S42, the heuristic initialization based on the maximum completion time is specifically as follows: For processes that can be processed at multiple workstations, the strategy for allocating workstations is as follows: First, the load of all workstations is 0 before the processing order is started. At this time, all the workstation numbers of the corresponding machine type are obtained from the workstation machine type array W according to the machine type required for processing the process, and the workstation with a high proficiency coefficient for processing the process is selected according to the worker proficiency coefficient matrix P. After that, each process that uses the machine type selects an unselected workstation according to the proficiency coefficient until all the workstations of the machine type are selected; After all the stations of the same machine type are selected, if there is a subsequent process that needs to use this type of machine for processing, the load and proficiency evaluation function of all stations in this machine type is calculated first. The calculation formula is as follows: Among them, t i,k The time of the i-th process in station k is used to represent the load, p i,k represents the proficiency coefficient corresponding to the i-th process in station k, n k represents the number of processes processed at station k. In this expression, if the load t of station k i,k The larger the value, the higher the evaluation score. k The smaller the proficiency coefficient p i,k The higher the score, the higher the evaluation score. k The larger the value is, the statistics of all the processes being processed in station k are used to calculate the load and proficiency evaluation score of station k. For all workstations of the same machine type, the evaluation score is E k Sort, each time you select a workstation based on machine type, select the largest E k The corresponding workstations are those with lighter loads and higher proficiency coefficients.

11. The method for scheduling clothing hanging production considering transportation distance according to claim 8, characterized in that: In step S45, the algorithm optimization stage includes the early and late optimization stages; in the early optimization stage, the three optimization goals of completion time, workstation load, and total transportation distance are sorted according to the optimization objective function values, and the population individuals are grouped into three intervals: high, medium, and low; in the late optimization stage, the three optimization goals of completion time, total transportation distance, and workstation load are respectively selected, and popSize / 3 optimal individuals are grouped.

12. The method for scheduling clothing hanging production considering transportation distance according to claim 8, characterized in that: In step S46, the similarity calculation formula is: S(i,j)=0.5S order +0.5S work ; in, n is the total length of the code, n s It is the number of characters that are exactly the same at the same position in the process level code; Count is the number of stations assigned to the same station under the same order process in the station level code.

13. The method for scheduling clothing hanging production considering transportation distance according to claim 12, characterized in that: The similarity matrix is: In the above formula, M[i][j]=S[i][j], M[i][j]=1 means that the similarity between the individual and itself is 1. The closer M[i][j] is to 1, the more similar the two individuals are. The closer it is to 0, the greater the difference between the two individuals.

14. The method for scheduling clothing hanging production considering transportation distance according to claim 13, characterized in that: The voting mechanism is as follows: Set a similarity threshold σ. If the similarity value S(i, j) between individuals i and j exceeds the set threshold σ, the two individuals are considered similar. According to the similarity matrix, each individual is voted. The voting method is to calculate the number of similar individuals between each individual and other individuals. The calculation formula is as follows: Among them, V i represents the voting score of an individual, ρ is a representation function, which is 1 when S[i][j]>σ, otherwise it is 0; If the voting score of individual i is V i If it exceeds 0.5popSize, that is, more than half of the individuals are similar to it, then individual i is considered to be a highly similar individual and part of the local optimal population; the number of highly similar individuals in the entire population is counted. When the number of highly similar individuals accounts for 60%-80% of the population, it is considered that the population has fallen into the local optimum and needs to be mutated. When the proportion is less than 60% or greater than 80% and the average improvement in fitness value is less than 1%, no mutation operation is required; The fitness improvement calculation formula is as follows: Among them, F n (i) represents the fitness value of the i-th individual in the n-th iteration.

15. A clothing hanging production scheduling system considering transportation distance, characterized in that: include: An acquisition module, which is used to acquire information of a clothing hanging production workshop, wherein the information of the clothing hanging production workshop includes clothing production information, clothing hanging production logs, and clothing hanging workshop layout; A processing module, which is used to process the information of the clothing hanging production workshop to obtain the worker proficiency coefficient matrix and the workstation distance matrix of the hanging workshop; A construction module is used to establish a mathematical model for clothing hanging production scheduling taking into account transportation distance by constructing an optimization objective function, adding a penalty mechanism, and defining constraints; A solution module, which is used to introduce an improved non-dominated sorting genetic algorithm to solve the mathematical model of clothing hanging production scheduling, so as to obtain a solution for clothing hanging production scheduling; Among them, the information of the clothing hanging production workshop is processed, including: According to the hanger flow history data obtained from the clothing hanging production workshop information, the proficiency coefficients of the designated workers at different workstations in the processing procedures are calculated, and the worker proficiency coefficient matrix is ​​statistically established. The expression is: In the above formula, p ijk Indicates that workstation k processes order O i The proficiency coefficient of the jth process is expressed as: In the above formula, α is the efficiency weight factor, β is the quality weight factor, and if α+β=1, the enterprise can determine it according to production strategy and market demand, and make corresponding adjustments according to the efficiency and quality requirements of different customers and orders; t ijk Indicates that workstation k processes order O i The actual processing time of the jth process is calculated by collecting the historical data of hanger flow and counting the time difference between the hanger entering and leaving the workstation k during the process; t ij Indicates processing order O i The standard processing time of the jth process; N i For order O i The total production quantity, and N r is the number of reworks required when processing this process at station k; t ijk The calculation formula is: In the above formula, t 进入 ,t 离开 They are processing orders O at station k respectively i The time for each hanger to enter and leave station k during the j-th process; According to the distance between workstations in the obtained clothing hanging production workshop information, the workstation distance matrix is ​​established using the adjacency matrix method, and the expression is: In the above formula, d is the distance between adjacent workstations on the same side, and d* is the distance between adjacent workstations in the U-bend of clothing hanging. The above matrix assumes that there are 4 workstations in the hanging workshop, and workstations 1, 2, 3, and 4 are on the same side. Therefore, the distance from workstation 1 to workstation 3 is d+d*. Due to the unidirectional nature of transportation, workstation 2 to workstation 1 needs to detour to workstations 3 and 4 to reach the destination, and the distance is d+2d*.

Images