Production order scheduling method in distributed manufacturing mode
By using triangular fuzzy numbers and GA-VNS algorithms in distributed manufacturing mode, a multi-objective optimization scheduling model is constructed, which solves the problems of factory qualification restrictions, transportation time and customer satisfaction, and provides reasonable scheduling for the production of new energy vehicle battery Decision-making improves productivity and cost-effectiveness.
Patent Information
- Application Number
- CN202411474698.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-22
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-10-22
AI Technical Summary
In the distributed manufacturing mode, it is difficult for the existing technology to effectively comprehensively consider the qualification restrictions, product transportation time and customer satisfaction issues of each factory, and it is impossible to apply the actual production needs of new energy vehicle batteries.
Triangular fuzzy numbers are used to represent the transportation time and expiration date of the order, and a multi-objective optimization scheduling model is built, combined with NSGA-II and VNS algorithms, and the Pareto frontier surface is obtained to obtain the optimal production order scheduling scheme.
Taking into account factory qualification restrictions, fuzzy transportation time and customer satisfaction, a method is provided to make reasonable scheduling decisions for new energy vehicle companies within a limited time, which improves the overall production capacity of the manufacturing system and reduces production costs.
Smart Images

Figure CN118982219A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of production order scheduling, and in particular to a production order scheduling method, system, storage medium and electronic equipment in a distributed manufacturing mode. Background Art
[0002] The rapid expansion of the energy vehicle industry has led to a surge in demand for batteries, prompting battery manufacturers to adopt a distributed manufacturing strategy. In this context, the production order scheduling problem of new energy vehicle batteries under the distributed manufacturing model has become a research hotspot.
[0003] In the related art, patent CN115130789A discloses an intelligent scheduling method for distributed manufacturing enterprises based on an improved gray wolf optimization algorithm. It takes into account uncertain factors such as machine failure and raw material shortage in the production process, defines the process time of the workpiece as a triangular fuzzy number, constructs a mathematical model with the optimization goal of minimizing the maximum fuzzy completion time, and designs a simple and efficient improved gray wolf optimization algorithm IGWO to solve the scheduling model. A three-dimensional encoding and decoding method based on processes, workshops and processing equipment is adopted for distributed workshops, and the quality of the initialization population is improved through a hybrid initialization strategy. The population position update formula is adjusted using an update mechanism based on nonlinear control parameters and adaptive inertia weights, and a local search strategy based on variable neighborhood search is adopted.
[0004] Although the above scheme takes into account uncertain factors such as machine failure and lack of raw materials during the production and processing, when considering realistic factors such as the unique qualification restrictions of each factory, the more easily affected product transportation time, and customer satisfaction issues, it is obvious that the scheduling scheme provided by the above patent is not well suited for the actual production of new energy vehicle batteries. Summary of the invention
[0005] 1. Technical issues to be resolved In view of the shortcomings of the prior art, the present invention provides a production order scheduling method, system, storage medium and electronic device under a distributed manufacturing model, which solves the technical problem that the prior art cannot be applied to the actual production of new energy vehicle batteries when considering realistic factors such as the qualification restrictions unique to each factory, the more easily affected product transportation time and customer satisfaction issues.
[0006] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: A production order scheduling method in a distributed manufacturing mode, comprising: Get a set of orders and a set of factories; each order includes only one product type, and each factory has only one production line for producing a specific type of product; Triangular fuzzy numbers are used to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction; Based on the order set and the factory set, a multi-objective optimization scheduling model is constructed with the goal of minimizing cost and maximizing customer satisfaction; The GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm is used to solve the optimization scheduling model and obtain the Pareto frontier to obtain the optimal production order scheduling solution.
[0007] Preferably, the method uses triangular fuzzy numbers to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction; including: Using triangular fuzzy numbers to represent order O i The fuzzy transportation time from factory j to customer i is denoted as ;in, , , Respectively represent order O i The shortest shipping time, the most likely shipping time, and the longest shipping time; Based on fuzzy transport time , get order O i The fuzzy arrival time ; Among them, once the order is processed, it is shipped to the customer immediately. ; , , Respectively represent order O i The shortest fuzzy arrival time, the most likely fuzzy arrival time and the longest fuzzy arrival time; C i Indicates order O i The completion time; m represents the number of factories; x ij is a 0-1 variable. If the order is O i If the factory produces in factory j, it takes 1, otherwise it takes 0; Using triangular fuzzy numbers to represent order O i The fuzzy expiration date is denoted as ;in, , , They represent the lower bound of customer i’s expected arrival time, the most expected arrival time, and the upper bound of the expected arrival time respectively; Based on fuzzy transport time and fuzzy expiration dates , obtain customer i’s satisfaction ; Wherein, area represents the area function and n represents the order quantity.
[0008] Preferably, the optimization scheduling model includes: Among them, the objective (1) represents minimizing the cost; Min represents minimizing the function, pc ij Indicates order O i Processing cost at plant j, tc ij Indicates order O i The determined transportation cost from plant j to customer i; Objective (2) represents maximizing customer satisfaction; Max represents the maximization function; Constraint (3) ensures that each order is assigned to one and only one factory; Constraint (4) limits orders to those plants that are capable of processing them; q ij is a 0-1 variable. If factory j has production order O i If the qualification is 1, it takes 1, otherwise it takes 0; Constraint (5) stipulates that each factory can have at most one order assigned as the first one to be processed, and some factories may have no orders assigned; y 0jk is a 0-1 variable, which takes the value 1 if the initial order is produced before order k in factory j, and takes the value 0 otherwise; Constraint (6) ensures that each order has a unique previous order, or the initial order in the factory queue; y ijk is a 0-1 variable. If order O is in factory j i In order O k If it is produced before, it takes 1, otherwise it takes 0; x kj is a 0-1 variable. If the order is O k If the factory produces in factory j, it takes 1, otherwise it takes 0; Constraint (7) states that every order has a successor order or is the last order in the factory queue; Constraint (8) excludes order O in factory j i It is Order O k The previous order of k The possibility of the next order; y kji is a 0-1 variable. If order O is in factory j k In order O i If it was produced before, it takes 1, otherwise it takes 0; Constraints (9) to (10) jointly determine the processing time of each order; C k Indicates order O k The completion time, M represents a large integer, p ij Indicates order O i processing time in plant j; Constraint (11) expresses the explicit decision variable x ij ,y ijk Is binary.
[0009] Preferably, the GA-VNS algorithm introduces an infinite archive set AS for dynamically maintaining the Pareto optimal solution set in the entire iteration process; the GA-VNS algorithm using the integrated NSGA-II algorithm and the VNS algorithm to solve the optimization scheduling model and obtain the Pareto frontier surface includes: Step 101: Enter order set , Factory Collection ,q ij 、p ij ,pc ij , 、tc ij , ; Step 102: Initialize parameters, including the maximum number of iterations t max , population size pop, rejection coefficient RC, mutation parameter gamma; Step 103: Initialize the population , perform non-dominated sorting on all individuals to screen the Pareto optimal solution and put it into the infinite archive set AS; The individual coding rule means that the first row of gene sequences of the chromosome indicates that the factory is determined for each order, and the second row of gene sequences indicates that the orders of the same factory are produced in ascending order of processing order values; Step 104, set the number of iterations t=0; Step 105: Traverse and select P from P i , and randomly generate an integer j in the range [1,pop], and j≠i; P i and P j Perform crossover operation as a parent and update the population P and infinite archive set AS based on the rejection coefficient RC; Step 106: Traverse and select P from P i , and calculate the mutation rate MR based on the mutation parameter gamma, perform the mutation operation, and update the population P and the infinite archive set AS; Step 107, set k=1; Step 108: When k≤2, randomly select P from P s , for P s Execute NS k Neighborhood search ; Otherwise, go to step 109; Step 109, calculate the fitness value, if fit( )>fit(P s), let P s = , k = 1, and go to step 108; otherwise, let k = k + 1, and go to step 108; Step 110: Update the population P and the infinite archive set AS; Step 111: Let t = t + 1; Step 112: If t ≤ t max , go to step 105, otherwise output the infinite archive set AS, and draw the Pareto front based on the infinite archive set AS.
[0010] Preferably, the specific steps of step 105 include: Step 1050: Let i = 1; Step 1051: Select P from P i , and randomly generate an integer j within the range of [1, pop], and j ≠ i; Step 1052: For the parent P i and P j , exchange the factory and processing order of the random segments to obtain the offspring C1 and C2; Step 1053: Design a fitness function based on the objective function of the optimization scheduling model: where fit(P x ) represents the fitness value of the individual P x , f1(P x ), f2(P x ) respectively represent the cost and customer satisfaction of the solution corresponding to the individual P x ; f1_max and f1_min respectively represent the maximum cost and minimum cost in the current population; f2_max and f2_min respectively represent the maximum satisfaction and minimum satisfaction in the current population; Step 1054: Generate a random number ∝ in (0, 1). If ∝ < fit(C1) ∙ RC, use the offspring C1 to replace the parent P i , if ∝ < fit(C2) ∙ RC, use the offspring C2 to replace the parent P j ; Step 1055: Let i = i + 1; Step 1056: If i ≤ pop, go to step 1051; otherwise end the crossover operation and update the infinite archive set AS.
[0011] Preferably, the specific steps of step 106 include: Step 1060: Let i = 1; Step 1061: Select P from P i , and obtain the current iteration number t, and calculate the mutation rate: Step 1062: Generate a random number ∝ of (0,1). If ∝>MR, then for individual P i , randomly change the factory and processing order of the order to obtain a new individual C; Step 1063, modify C according to factory qualification restrictions; Step 1064, let i=i+1; Step 1065: If i≤pop, go to step 1061; otherwise, end the mutation operation and update the infinite archive set AS.
[0012] Preferably, the following two neighborhood structures are designed: (1) Neighborhood structure 1 with the goal of optimizing cost, including: Step 1a: Input individual P s ; Step 1b, traverse the factories, select the order with the highest cost in each factory, assign it to the remaining factories, and select the factory with the lowest cost; Step 1c: Compare the fitness values and update P s ; (2) Neighborhood structure 2 with the goal of optimizing customer satisfaction, where for a triangular fuzzy number =(A L , A C ,A R ),use =( A L + 2A C + A R ) / 4 represents the distance to the y-axis, A L , A C , A R They represent the lower bound, most likely value and upper bound of the variable A respectively; The neighborhood structure 2 comprises: Step 2a: Input individual P s ; Step 2b: Traverse the factories and randomly select order O in each factory x ,calculate , x represents order O x The sequence of processing in the factory; Step 2c: If , order O x Exchange the processing order with the subsequent order, otherwise, order O x Exchange the processing order with the previous order to find the optimal processing order; among them, Indicates order O x The fuzzy arrival time of Indicates order O x The fuzzy expiration date of Step 2d: Compare the fitness values and update P s .
[0013] A production order scheduling system under a distributed manufacturing mode, comprising: A data acquisition module, used to acquire an order set and a factory set; each order includes only one product type, and each factory has only one production line for producing a specific type of product; The satisfaction building module is used to construct customer satisfaction by using triangular fuzzy numbers to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order; A model building module, used to build a multi-objective optimization scheduling model based on the order set and the factory set with the goal of minimizing cost and maximizing customer satisfaction; The solution solving module is used to solve the optimization scheduling model by using the GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm, obtain the Pareto frontier surface, and obtain the optimal production order scheduling solution.
[0014] A storage medium stores a computer program for production order scheduling in a distributed manufacturing mode, wherein the computer program enables a computer to execute the production order scheduling method in the distributed manufacturing mode as described above.
[0015] An electronic device, comprising: One or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the programs include a method for executing the production order scheduling method in the distributed manufacturing mode as described above.
[0016] (III) Beneficial effects The present invention provides a method, system, storage medium and electronic device for scheduling production orders in a distributed manufacturing mode. Compared with the prior art, the present invention has the following beneficial effects: In the present invention, firstly, the order set and the factory set are obtained; secondly, triangular fuzzy numbers are used to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order, so as to construct customer satisfaction; again, based on the order set and the factory set, a multi-objective optimization scheduling model is constructed with the goal of minimizing cost and maximizing customer satisfaction; finally, the GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm is used to solve the optimization scheduling model, obtain the Pareto frontier surface, and obtain the optimal production order scheduling plan. By incorporating factory qualification restrictions, fuzzy transportation time and customer satisfaction into the model framework, the problem is closer to reality. At the same time, an optimization algorithm combining genetic algorithm and variable neighborhood search algorithm is designed, which can make reasonable scheduling decisions for new energy vehicle companies within a limited time. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0018] Figure 1 A block diagram of a production order scheduling method in a distributed manufacturing mode provided by an embodiment of the present invention; Figure 2 An example diagram of a chromosome provided by an embodiment of the present invention; Figure 3 A satisfaction diagram provided by an embodiment of the present invention; Figure 4 An example diagram of a Pareto front surface provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0020] The embodiments of the present application provide a production order scheduling method, system, storage medium and electronic device in a distributed manufacturing mode, thereby solving the technical problem that the existing technology cannot be applied to the actual production of new energy vehicle batteries when considering realistic factors such as the qualification restrictions unique to each factory, the more easily affected product transportation time and customer satisfaction issues.
[0021] The technical solution in the embodiment of the present application is to solve the above technical problems, and the overall idea is as follows: This paper aims to solve the distributed scheduling problem of production orders and factories. In order to make the problem more realistic, factory qualification restrictions, fuzzy transportation time and customer satisfaction are incorporated into the model framework to more fully describe the problem. This integration not only enriches existing research, but also brings the research closer to real-world applications, making it more practical and relevant.
[0022] At the same time, according to the characteristics of the new energy vehicle battery production mode, a corresponding heuristic algorithm is designed, and a combined optimization strategy of genetic algorithm and variable neighborhood search is proposed, which can make reasonable scheduling decisions for enterprises within a limited time, or provide decision support for enterprise decision makers, thereby improving the overall production capacity of the manufacturing system and reducing production costs.
[0023] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0024] Embodiment 1: like Figure 1 As shown, an embodiment of the present invention provides a production order scheduling method in a distributed manufacturing mode, comprising: S1. Obtain an order set and a factory set; each order includes only one product type, and each factory has only one production line for producing a specific type of product; S2, using triangular fuzzy numbers to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction; S3. Based on the order set and the factory set, a multi-objective optimization scheduling model is constructed with the goal of minimizing cost and maximizing customer satisfaction; S4. A GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm is used to solve the optimization scheduling model and obtain the Pareto frontier to obtain the optimal production order scheduling solution.
[0025] The embodiment of the present invention makes the problem closer to reality by incorporating factory qualification restrictions, fuzzy transportation time and customer satisfaction into the model framework. At the same time, an optimization algorithm combining a genetic algorithm and a variable neighborhood search algorithm is designed, which can make reasonable scheduling decisions for new energy vehicle companies within a limited time.
[0026] Next, we will introduce the steps of the above scheme in detail: In step S1, an order set and a factory set are obtained; each order includes only one product type, and each factory has only one production line for producing a specific type of product.
[0027] The embodiment of the present invention studies the distributed scheduling problem of production orders and factories. The description of this problem can be expressed as follows: there is a set of orders, represented as , which needs to be processed in a set of factories, represented as . The only order O from customer i i There is a fuzzy expiration date, denoted by Each order contains only one product type, and there is no predefined priority between these orders. Each factory has a unique production line that can only produce a specific range of product types. The processing time of an order in a factory is denoted by p ij , the corresponding processing cost is denoted as pc ij Once the orders are completed, they are shipped directly from their respective factories to the intended customers.
[0028] It should be noted that setting each order to include only one product type and each factory to have only one production line for producing a specific type of product is consistent with the distributed manufacturing strategy adopted by battery manufacturers, while taking into account the unique qualification constraints of each factory (i.e., the ability of different factories to produce specific types of orders).
[0029] In order to facilitate the smooth construction of the subsequent model, the variables and symbols shown in Table 1 are given.
[0030] Table 1 In step S2, triangular fuzzy numbers are used to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction.
[0031] In this step, the influence of fuzzy shipping time and fuzzy expiration date on customer satisfaction is discussed based on fuzzy set theory, as follows: Using triangular fuzzy numbers to represent order O i The fuzzy transportation time from factory j to customer i is denoted as ;in, , , Respectively represent order O i The shortest shipping time, the most likely shipping time, and the longest shipping time; Based on fuzzy transport time , get order O i The fuzzy arrival time ; Among them, once the order is processed, it is shipped to the customer immediately. ; , , Respectively represent order O iThe shortest fuzzy arrival time, the most likely fuzzy arrival time, and the longest fuzzy arrival time; Using triangular fuzzy numbers to represent order O i The fuzzy expiration date is denoted as ;in, , , They represent the lower bound of customer i’s expected arrival time, the most expected arrival time, and the upper bound of the expected arrival time respectively; Based on fuzzy transport time and fuzzy expiration dates ,like Figure 3 As shown, obtain the satisfaction of customer i ; Wherein, area represents the area function.
[0032] In step S3, based on the order set and the factory set, a multi-objective optimization scheduling model is constructed with the goal of minimizing cost and maximizing customer satisfaction. The optimization scheduling model constructed in this step has two goals: first, to minimize the total cost, which includes the production cost and transportation cost of all orders; second, to maximize the overall customer satisfaction level, where the satisfaction of a single customer is solved based on the content in step 2.
[0033] Specifically, the optimization scheduling model includes: Among them, objective (1) represents minimizing cost; Min represents minimizing function; Objective (2) represents maximizing customer satisfaction; Max represents the maximization function; Constraint (3) ensures that each order is assigned to one and only one factory; Constraint (4) limits orders to those plants that are capable of processing them; Constraint (5) stipulates that each plant can have at most one order assigned as the first to be processed, while some plants may have no orders assigned; Constraint (6) ensures that each order has a unique previous order, or the initial order in the factory queue; Constraint (7) states that every order has a successor order or is the last order in the factory queue; Constraint (8) excludes order O in factory j i It is Order O k The previous order of k The possibility of a subsequent order; Constraints (9) to (10) jointly determine the processing time of each order; The constraint (11) represents the explicit decision variables x ij , y ijk is binary.
[0034] In step S4, the GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm is used to solve the optimization scheduling model, obtain the Pareto front, so as to obtain the optimal production order scheduling scheme.
[0035] In this step, the NSGA-II algorithm and the VNS algorithm are integrated. This integration aims to utilize the exploratory advantages of NSGA-II and the local search ability of VNS. By strategically increasing the mutation rate of the genetic algorithm, local optima are effectively avoided, thus promoting the rapid discovery of the optimal solution through neighborhood search. To further simplify the search process, an infinite archive set AS is also introduced, which dynamically maintains the Pareto optimal solution set throughout the iteration process Specifically, this step includes:[[]] Step 101, input the order set , the factory set , q ij , p ij , pc ij , , tc ij , ; Step 102, initialize the parameters, including the maximum number of iterations t max , the population size pop, the rejection coefficient RC, and the mutation parameter gamma; Step 103, initialize the population , perform non-dominated sorting on all individuals to screen the Pareto optimal solutions and put them into the infinite archive set AS; Among them, the encoding rule of an individual means that the first row gene sequence of the chromosome represents determining the factory for each order respectively, and the second row gene sequence represents producing the orders of the same factory in ascending order of the processing order value.
[0036] Specifically, each order has a factory FN and a processing order OS. After determining the factory, the orders of the same factory are produced in ascending order of the OS value. For clearer visualization Figure 2 , an example with 8 orders and 2 factories is given: at the FN level, it can be obtained that orders O2, O4, and O5 are assigned to factory 1, and the remaining orders are assigned to factory 2; at the OS level, it can be obtained that in factory 1, OS(O4) < OS(O2) < OS(O5), and the orders are produced in the order of O4 → O2 → O5; similarly, in factory 2, in the order of O3 → O7 → O6 → O1 → O8.
[0037] It should be understood that non - dominated sorting is performed on all individuals in the population in this step, and then the Pareto optimal solutions are screened, which is the existing technology (the basic content in the NSGA - II algorithm), and will not be elaborated here.
[0038] Step 104: Let the iteration number t = 0; Step 105: Traverse and select P from P i , and randomly generate an integer j in the range of [1, pop], and j ≠ i; Take P i and P j as parents to perform the crossover operation, and update the population P and the infinite archive set AS based on the rejection coefficient RC. Specifically, it includes: Step 1050: Let i = 1; Step 1051: Select P from P i , and randomly generate an integer j in the range of [1, pop], and j ≠ i; Step 1052: For the parents P i and P j , exchange the factories and processing orders of the random segments to obtain the offspring C1 and C2; Step 1053: Design a fitness function based on the objective function of the optimization scheduling model: where, fit(P x ) represents the fitness value of the individual P x , f1(P x ), f2(P x ) respectively represent the cost and customer satisfaction of the solution corresponding to the individual P x ; f1_max and f1_min respectively represent the maximum cost and minimum cost in the current population; f2_max and f2_min respectively represent the maximum satisfaction and minimum satisfaction in the current population; Step 1054: Generate a random number ∝ in (0, 1). If ∝ < fit(C1) ∙ RC, use the offspring C1 to replace the parent P i , if ∝ < fit(C2) ∙ RC, use the offspring C2 to replace the parent P j ; Step 1055: Let i = i + 1; Step 1056: If i ≤ pop, transfer to Step 1051; otherwise, end the crossover operation and update the infinite archive set AS.
[0039] Step 106: Traverse and select P from P i , and calculate the mutation rate MR based on the mutation parameter gamma, perform the mutation operation, and update the population P and the infinite archive set AS. Specifically, it includes: Step 1060: Let i = 1; Step 1061: Select P from P i , and get the current iteration number t, and calculate the mutation rate: Step 1062: Generate a random number ∝ of (0,1). If ∝>MR, then for individual P i , randomly change the factory and processing order of the order to obtain a new individual C; Step 1063, modify C according to factory qualification restrictions; Step 1064, let i=i+1; Step 1065: If i≤pop, go to step 1061; otherwise, end the mutation operation and update the infinite archive set AS.
[0040] Step 107: Set k=1.
[0041] Step 108: When k≤2, randomly select P from P s , for P s Execute NS k Neighborhood search ; Otherwise, go to step 109.
[0042] Exemplarily, the following two neighborhood structures are designed here: (1) Neighborhood structure 1 with the goal of optimizing cost, including: Step 1a: Input individual P s ; Step 1b, traverse the factories, select the order with the highest cost in each factory, assign it to the remaining factories, and select the factory with the lowest cost; Step 1c: Compare the fitness values and update P s ; (2) Neighborhood structure 2 with the goal of optimizing customer satisfaction, where for a triangular fuzzy number =(A L , A C ,A R ),use =( A L + 2A C + A R ) / 4 represents the distance to the y-axis, A L , A C , A R They represent the lower bound, most likely value and upper bound of the variable A respectively; The neighborhood structure 2 comprises: Step 2a: Input individual P s ; Step 2b: Traverse the factories and randomly select order O in each factory x ,calculate , x represents order O x The sequence of processing in the factory; Step 2c: If , order O x Exchange the processing order with the subsequent order, otherwise, order O x Exchange the processing order with the previous order to find the optimal processing order; among them, Indicates order O x The fuzzy arrival time of Indicates order O x The fuzzy expiration date of Step 2d: Compare the fitness values and update P s .
[0043] Step 109, calculate the fitness value, if fit( )>fit(P s ), let P s = , k=1, and go to step 108; otherwise, set k=k+1 and go to step 108.
[0044] Step 110: Update the population P and the infinite archive set AS.
[0045] Step 111, let t=t+1.
[0046] Step 112: If t≤t max , go to step 105, otherwise output the infinite archive set AS, and draw the Pareto front surface based on the infinite archive set AS.
[0047] Step 113: Based on the Pareto frontier, the optimal solution can be obtained according to actual needs and preferences, and the optimal production order scheduling solution can be obtained by performing a decoding operation.
[0048] It is understandable that in multi-objective optimization problems, there is often no single optimal solution, but rather multiple conflicting optimal solutions. Each point on the Pareto front represents a possible solution that achieves the best trade-off between multiple objectives. It can provide decision makers with comprehensive information and multiple options, allowing them to fully understand the various possibilities and choose the appropriate solution based on actual needs and preferences.
[0049] Exemplarily, during implementation, the experimental parameters are randomly generated, specifically: Parameters pc can be set ij and tc ijThe value range of is set between [1,100], and p ij Varies in the interval [1,50]. , Randomly selected from the interval [1,20], and and Respectively in the interval [ , 30] and the interval [ , 50] internal changes. The boundaries are [10, 50], [ , 100],[ , 150]. Set RC to 0.8 and the population size pop to 50.
[0050] In addition, by adjusting gamma and t max A sensitivity analysis was performed to determine the optimal values of gamma=0.05 and t max =50, so the Pareto frontier is Figure 4 shown.
[0051] So far, the embodiment of the present invention has completed the entire process of the production order scheduling method in the distributed manufacturing mode.
[0052] Embodiment 2: An embodiment of the present invention provides a production order scheduling system in a distributed manufacturing mode, including: A data acquisition module, used to acquire an order set and a factory set; each order includes only one product type, and each factory has only one production line for producing a specific type of product; The satisfaction building module is used to construct customer satisfaction by using triangular fuzzy numbers to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order; A model building module, used to build a multi-objective optimization scheduling model based on the order set and the factory set with the goal of minimizing cost and maximizing customer satisfaction; The solution solving module is used to solve the optimization scheduling model by using the GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm, obtain the Pareto frontier surface, and obtain the optimal production order scheduling solution.
[0053] Embodiment 3: An embodiment of the present invention provides a storage medium storing a computer program for production order scheduling in a distributed manufacturing mode, wherein the computer program enables a computer to execute the production order scheduling method in a distributed manufacturing mode as described in Example 1.
[0054] Embodiment 4: An embodiment of the present invention provides an electronic device, including: One or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the programs include a method for executing a production order scheduling method in a distributed manufacturing mode as described in Example 1.
[0055] It can be understood that the production order scheduling system, storage medium and electronic device under the distributed manufacturing mode provided by the embodiments of the present invention correspond to the production order scheduling method under the distributed manufacturing mode provided by the embodiments of the present invention, and the explanations, examples and beneficial effects of the relevant contents can refer to the corresponding parts of the production order scheduling method under the distributed manufacturing mode, and will not be repeated here.
[0056] In summary, compared with the prior art, the present invention has the following beneficial effects: 1. The embodiment of the present invention solves the distributed scheduling problem of production orders and factories. In order to make the problem more realistic, factory qualification restrictions, fuzzy transportation time and customer satisfaction are incorporated into the model framework to provide a more complete description of the problem. This integration not only enriches existing research, but also brings the research closer to real-world applications, making it more practical and relevant.
[0057] 2. The distributed scheduling optimization method constructed in the embodiment of the present invention is based on a non-dominated genetic algorithm and a variable neighborhood search algorithm, and introduces an infinite capacity storage set AS to store the Pareto solution set, so as to store more Pareto optimal solutions and obtain a better Pareto frontier; and it can record all non-dominated solutions generated during the iteration process to meet diverse needs.
[0058] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.
[0059] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A production order scheduling method under distributed manufacturing mode, It is characterized in that include: Get the order collection and factory collection; Each order includes only one type of product, and each factory has only one production line for producing a specific type of product; Triangular fuzzy numbers are used to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction; Based on the order set and the factory set, a multi-objective optimization scheduling model is constructed with the goal of minimizing cost and maximizing customer satisfaction; The GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm is used to solve the optimization scheduling model and obtain the Pareto frontier to obtain the optimal production order scheduling solution.
2. The method for scheduling production orders in a distributed manufacturing mode according to claim 1, It is characterized in that The triangular fuzzy numbers are used to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order to construct customer satisfaction; including: Using triangular fuzzy numbers to represent order O i The fuzzy transportation time from factory j to customer i is denoted as ;in, , , Respectively represent order O i The shortest shipping time, the most likely shipping time, and the longest shipping time; Based on fuzzy transport time , get order O i The fuzzy arrival time ; Among them, once the order is processed, it is shipped to the customer immediately. ; , , Respectively represent order O i The shortest fuzzy arrival time, the most likely fuzzy arrival time and the longest fuzzy arrival time; C i Indicates order O i The completion time; m represents the number of factories; x ij is a 0-1 variable. If the order is O i If the factory produces in factory j, it takes 1, otherwise it takes 0; Using triangular fuzzy numbers to represent order O i The fuzzy expiration date is denoted as ;in, , , They represent the lower bound of customer i’s expected arrival time, the most expected arrival time, and the upper bound of the expected arrival time respectively; Based on fuzzy transport time and fuzzy expiration dates , obtain customer i’s satisfaction ; Wherein, area represents the area function and n represents the order quantity.
3. The method for scheduling production orders in a distributed manufacturing mode as claimed in claim 2, It is characterized in that The optimization scheduling model includes: Among them, the objective (1) represents minimizing the cost; Min represents minimizing the function, pc ij Indicates order O i Processing cost at plant j, tc ij Indicates order O i The determined transportation cost from plant j to customer i; Objective (2) represents maximizing customer satisfaction; Max represents the maximization function; Constraint (3) ensures that each order is assigned to one and only one factory; Constraint (4) limits orders to those plants that are capable of processing them; q ij is a 0-1 variable. If factory j has production order O i If the qualification is 1, it takes 1, otherwise it takes 0; Constraint (5) stipulates that each factory can have at most one order assigned as the first one to be processed, and some factories may have no orders assigned; y 0jk is a 0-1 variable, which takes the value 1 if the initial order is produced before order k in factory j, and takes the value 0 otherwise; Constraint (6) ensures that each order has a unique previous order, or the initial order in the factory queue; y ijk is a 0-1 variable. If order O is in factory j i In order O k If it is produced before, it takes 1, otherwise it takes 0; x kj is a 0-1 variable. If the order is O k If the factory produces in factory j, it takes 1, otherwise it takes 0; Constraint (7) states that every order has a successor order or is the last order in the factory queue; Constraint (8) excludes order O in factory j i It is Order O k The previous order of k The possibility of the next order; y kji is a 0-1 variable. If order O is in factory j k In order O i If it was produced before, it takes 1, otherwise it takes 0; Constraints (9) to (10) jointly determine the processing time of each order; C k Indicates order O k The completion time, M represents a large integer, p ij Indicates order O i processing time in plant j; Constraint (11) expresses the explicit decision variable x ij ,y ijk Is binary.
4. The method for scheduling production orders in a distributed manufacturing mode as claimed in claim 3, It is characterized in that The GA-VNS algorithm introduces an infinite archive set AS for dynamically maintaining the Pareto optimal solution set in the entire iteration process; the GA-VNS algorithm using the integrated NSGA-II algorithm and the VNS algorithm solves the optimization scheduling model to obtain the Pareto frontier; including: Step 101: Enter order set , Factory Collection ,q ij 、p ij ,pc ij , 、tc ij , ; Step 102: Initialize parameters, including the maximum number of iterations t max , population size pop, rejection coefficient RC, mutation parameter gamma; Step 103: Initialize the population , perform non-dominated sorting on all individuals to screen the Pareto optimal solution and put it into the infinite archive set AS; The individual coding rule means that the first row of gene sequences of the chromosome indicates that the factory is determined for each order, and the second row of gene sequences indicates that the orders of the same factory are produced in ascending order of processing order values; Step 104, set the number of iterations t=0; Step 105: Traverse and select P from P i , and randomly generate an integer j in the range [1,pop], and j≠i; P i and P j Perform crossover operation as a parent and update the population P and infinite archive set AS based on the rejection coefficient RC; Step 106: Traverse and select P from P i , and calculate the mutation rate MR based on the mutation parameter gamma, perform the mutation operation, and update the population P and the infinite archive set AS; Step 107, set k=1; Step 108: When k≤2, randomly select P from P s , for P s Execute NS k Neighborhood search ; Otherwise, go to step 109; Step 109, calculate the fitness value, if fit( )>fit(P s ), let P s = , k=1, and go to step 108; otherwise, set k=k+1, and go to step 108; Step 110, updating the population P and the infinite archive set AS; Step 111, let t=t+1; Step 112: If t≤t max , go to step 105, otherwise output the infinite archive set AS, and draw the Pareto front surface based on the infinite archive set AS.
5. The method for scheduling production orders in a distributed manufacturing mode as claimed in claim 4, It is characterized in that The step 105 specifically includes: Step 1050, let i=1; Step 1051: Select P from P i , and randomly generate an integer j in the range [1,pop], and j≠i; Step 1052: for parent P i and P j , swapping the factories and processing order of random fragments to obtain progeny C 1 and C 2 ; Step 1053: Design a fitness function based on the objective function of the optimization scheduling model: Among them, fit(P x ) represents individual P x The fitness value, f 1 (P x ), f 2 (P x ) represent the individual P x The cost of the corresponding solution and customer satisfaction; f 1 _max, f 1 _min represents the maximum cost and minimum cost in the current population; f 2 _max, f 2 _min represents the maximum satisfaction and minimum satisfaction in the current population; Step 1054, generate a random number ∝ of (0,1), if ∝ <fit(C 1 )∙RC, using descendant C 1 Replace parent P i , if ∝ <fit(C 2 )∙RC, using descendant C 2 Replace parent P j ; Step 1055, let i=i+1; Step 1056: If i≤pop, go to step 1051; otherwise, end the crossover operation and update the infinite archive set AS.
6. The method for scheduling production orders in a distributed manufacturing mode as claimed in claim 4, It is characterized in that The step 106 specifically includes: Step 1060, let i=1; Step 1061: Select P from P i , and get the current iteration number t, and calculate the mutation rate: Step 1062: Generate a random number ∝ of (0,1). If ∝>MR, then for individual P i , randomly change the factory and processing order of the order to obtain a new individual C; Step 1063, modify C according to factory qualification restrictions; Step 1064, let i=i+1; Step 1065: If i≤pop, go to step 1061; otherwise, end the mutation operation and update the infinite archive set AS.
7. The method for scheduling production orders in a distributed manufacturing mode as claimed in claim 4, It is characterized in that The following two neighborhood structures are designed: (1) Neighborhood structure 1 with the goal of optimizing cost, including: Step 1a: Input individual P s ; Step 1b, traverse the factories, select the order with the highest cost in each factory, assign it to the remaining factories, and select the factory with the lowest cost; Step 1c: Compare the fitness values and update P s ; (2) Neighborhood structure 2 with the goal of optimizing customer satisfaction, where for a triangular fuzzy number =(A L , A C , A R ),use =( A L + 2A C + A R ) / 4 represents the distance to the y-axis, A L , A C , A R They represent the lower bound, most likely value and upper bound of the variable A respectively; The neighborhood structure 2 comprises: Step 2a: Input individual P s ; Step 2b: Traverse the factories and randomly select order O in each factory x ,calculate , x represents order O x The sequence of processing in the factory; Step 2c: If , order O x Exchange the processing order with the subsequent order, otherwise, order O x Exchange the processing order with the previous order to find the optimal processing order; among them, Indicates order O x The fuzzy arrival time of Indicates order O x The fuzzy expiration date of Step 2d: Compare the fitness values and update P s .
8. A production order scheduling system under distributed manufacturing mode, It is characterized in that include: Data acquisition module, used to obtain order collection and factory collection; Each order includes only one type of product, and each factory has only one production line for producing a specific type of product; The satisfaction building module is used to construct customer satisfaction by using triangular fuzzy numbers to represent the fuzzy transportation time of the order from the factory to the customer and the fuzzy expiration date of the order; A model building module, used to build a multi-objective optimization scheduling model based on the order set and the factory set with the goal of minimizing cost and maximizing customer satisfaction; The solution solving module is used to solve the optimization scheduling model by using the GA-VNS algorithm integrating the NSGA-II algorithm and the VNS algorithm, obtain the Pareto frontier surface, and obtain the optimal production order scheduling solution.
9. A storage medium, It is characterized in that It stores a computer program for production order scheduling in a distributed manufacturing mode, wherein the computer program enables a computer to execute the production order scheduling method in a distributed manufacturing mode as described in any one of claims 1 to 7.
10. An electronic device, It is characterized in that include: one or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the programs include a method for executing the production order scheduling method under the distributed manufacturing mode as described in any one of claims 1 to 7.
Citation Information
Patent Citations
Distributed factory production distribution integrated scheduling method and system
CN114594744A
Multi-objective optimization method for multi-factory production scheduling under uncertain sea transportation time
CN115759646A
Multi-objective optimization method and system for master production plan of casting parallel workshops
US20210373888A1
Cited By
Equipment production scheduling multi-objective optimization method and system based on multi-chain collaboration
CN119761722A
Multi-chain coordination-based equipment production scheduling multi-objective optimization method and system
CN119761722B
Manufacturing system multi-objective decision-making method and system based on prediction model
CN121094607A
