Simulation-based distributed heterogeneous machining workshop order allocation method and system

Abstract

The invention relates to the technical field of intelligent optimization of discrete combination problems, in particular to a distributed heterogeneous machining workshop order allocation method and system based on simulation, workshops are accurately described by constructing simulation models of the machining workshops, and a hybrid genetic algorithm is combined with simulation to find order allocation schemes meeting requirements, so that the optimization of the machining workshops is realized. According to the method, a factory delivery date relaxation calculation mode is improved, maximization and minimum factory delivery date relaxation are taken as a target, a high-quality order allocation scheme is efficiently generated, whether each order can be completed in time or not is verified through production simulation of each machining workshop, an order transfer operation is designed to adjust overdue order allocation, and the order allocation efficiency is improved. Compared with a traditional method, the efficiency of processing heterogeneous workshops is higher, order allocation of simulation verification is more real and reliable, the resource utilization rate and the on-time completion rate can be improved in a complex distributed manufacturing scene, the order delay risk is reduced, and the method is suitable for the field of multi-factory collaborative production scheduling.

Description

Technical Field

[0001] This invention relates to the field of intelligent optimization technology for discrete combinatorial problems, specifically to a simulation-based method and system for allocating orders in a distributed heterogeneous machining workshop. Background Technology

[0002] In modern manufacturing, the problem of order allocation in distributed machining workshops is becoming increasingly complex. With the popularization of multi-factory collaborative production models, there are significant differences in equipment configuration, process flow and production constraints among workshops, making it difficult for traditional order allocation methods to efficiently adapt to actual production needs.

[0003] In existing technologies, common order allocation schemes are mostly based on rule-driven or traditional optimization algorithms, such as linear programming and metaheuristic algorithms. These methods usually assume that workshop resources are homogeneous and cannot take into account the differences in process capabilities between heterogeneous workshops. This leads to a disconnect between the allocation scheme and actual production, low resource utilization, and high risk of order completion delays. At the same time, existing technologies lack consideration for the dynamic performance of the workshop and effective verification mechanisms when allocating orders. They rely heavily on experience judgment or static evaluation models, making it difficult to dynamically simulate the actual production process of orders in heterogeneous workshops. Some workshop equipment may fail, and traditional methods cannot consider and model these failures. Although some methods have introduced simulation technology to assist scheduling, their simulation models are often oversimplified and cannot fully reflect the real-time production performance of the workshop. This results in a large deviation between the simulation results and the real scenario, making it difficult to support high-quality order allocation decisions. Therefore, a simulation-based distributed heterogeneous machining workshop order allocation method and system is proposed. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a simulation-based method and system for allocating orders in a distributed heterogeneous machining workshop.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a simulation-based distributed heterogeneous machining workshop order allocation method and system, wherein the order allocation method specifically includes the following steps:

[0006] The order allocation method specifically includes the following steps:

[0007] S1 constructs a discrete event simulation model based on the actual conditions of each heterogeneous machining workshop;

[0008] S2 uses a genetic algorithm based on single-layer coding to globally optimize the order allocation scheme. The genetic algorithm uses the minimum delivery slack value among all workshops as the objective function to determine the optimal workshop assignment for each order.

[0009] S3. According to the order allocation plan, each order is input into the discrete event simulation model of the workshop. In each workshop, the production sequence of the orders is determined based on the delivery date. The production execution process is simulated. After running the simulation, the delivery date of each order is obtained, and the simulation verification is realized.

[0010] S4. Based on the delivery dates of each order in the simulation results, determine whether the on-time completion rate of the orders meets the requirements. If the requirements are met, output the final order allocation plan. Otherwise, perform local optimization, execute the order reallocation operation, identify the order with the longest delay, and migrate the overdue order to the target workshop with processing capacity based on the status output by the simulation model of each workshop.

[0011] S5, Update the order allocation plan based on the order migration results;

[0012] S6. Repeat steps S3-S5 until the maximum number of iterations is met or the on-time completion rate of the order meets the requirements.

[0013] Preferably, step S1 specifically includes the following steps:

[0014] S11. Based on the equipment layout and resource list of each distributed machining workshop, create entity model objects, including processing equipment, buffer, tooling resources, worker pool, warehouse, loaders, assembly station, disassembly station, conveyor belt, AGV, RGV and overhead crane. Then define the logistics scheduling module and worker scheduling module and set the actual characteristics.

[0015] S12, Import process information, including process time, process sequence, resource requirements, worker requirements, transfer tools, batch size and product model for each product;

[0016] S13, edit the script in the discrete event simulation model to ensure that each product runs according to the actual process route and process time;

[0017] S14, collect the order information to be processed, simulate and verify the module, and put it into production in the following order: the order with the closer delivery date is put into production first. If the delivery dates of two orders are the same, the order with the smaller delivery date slack is put into production first.

[0018] The initial state of the discrete event simulation model is set according to the real-time status of the workshop.

[0019] Preferably, in step S2, the single-layer encoding method is as follows:

[0020] Each chromosome represents an order allocation scheme to a workshop. The chromosome length is equal to the total number of orders. The gene value of the i-th gene position represents the workshop number allocated to the i-th order. All gene values ​​are within the range of valid workshop numbers.

[0021] Preferably, in step S2, the objective function is specifically:

[0022] In step S2, the objective function is specifically:

[0023] For each order assigned to a particular workshop, the difference between the delivery date of that order and the fastest completion time in that workshop is the delivery slack. Specifically:

[0024] ;

[0025] in, This represents the delivery date of order n. This represents the fastest completion time required when only order n is processed in the current workshop;

[0026] The total lead time slack for each workshop is the sum of the lead time slacks of all orders allocated to that workshop. ,Right now:

[0027] ;

[0028] in, This represents the set of all orders assigned to the i-th workshop;

[0029] Initialize algorithm parameters and order information; the minimum factory delivery time slack is set to the fitness value. ,Right now;

[0030]

[0031] The objective function is to maximize the minimum total delivery slack across all workshops, i.e.:

[0032]

[0033] Where m represents the total number of workshops, which is used to improve the overall robustness of the system upon completion and to avoid excessive pressure on local workshops.

[0034] Preferably, in step S3, the simulation verification includes the following steps:

[0035] S31, simulates scheduling and execution according to the order process route and resource constraints in the simulation model, and outputs the actual completion time of each order;

[0036] S32, if an order's completion time is later than its delivery date, it is determined to be an overdue order, triggering the subsequent order reallocation mechanism;

[0037] S33. If the overall on-time completion rate reaches the preset threshold or the upper limit of the number of order reassignments, then the current order allocation plan is confirmed to be feasible.

[0038] Preferably, in step S4, the order reallocation operation includes the following steps:

[0039] S41, Identify overdue orders and their key bottleneck processes in the original workshop;

[0040] S42, based on the real-time status output by the simulation model of each workshop, evaluates three indicators of the candidate workshop: remaining effective capacity, current delivery slack, and compatibility with the process route of the order.

[0041] S43, select the workshop with the highest comprehensive evaluation score of the three indicators as the target workshop, execute the order migration and update the production plan of the relevant workshop;

[0042] S44, Rerun the simulation to verify the adjusted overall scheme.

[0043] A simulation-based distributed heterogeneous machining workshop order allocation system, comprising a simulation modeling module, an algorithm optimization module, a simulation verification module, an order migration module, and a scheduling and coordination engine;

[0044] The simulation modeling module is used to build and maintain discrete event simulation models for multiple heterogeneous machining workshops;

[0045] The algorithm optimization module uses a genetic algorithm to randomly generate and optimize the initial order allocation scheme.

[0046] The simulation verification module is used to simulate the production execution process and obtain the delivery dates for each order;

[0047] The order migration module is used to perform order reassignment operations;

[0048] The scheduling and coordination engine is used to update orders and input simulation models to obtain simulation results.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0050] 1. This invention combines optimization algorithms with production simulation. The genetic algorithm uses a fitness function with delivery slack as the core, avoiding the introduction of process scheduling and circumventing the problem of difficult process coding in distributed heterogeneous workshops. It quickly generates an initial allocation scheme and significantly reduces computational costs. By running the simulation model, orders are input into the discrete event simulation model according to the production order and the simulation is run to obtain the actual completion time of each order. The actual completion time is compared with the planned completion time to verify the feasibility of the order allocation scheme. Furthermore, the risk of overdue orders is reduced through a local reallocation strategy. The results are more realistic and significantly improve the overall scheduling efficiency.

[0051] 2. This invention introduces production simulation, which closely matches the actual production environment, reducing the risk of disconnect between planning and execution. The simulation model fully considers real-world disturbance factors such as equipment resources, workers, and malfunctions, and performs high-fidelity simulation of order completion time. The overdue order identification and migration mechanism based on simulation feedback effectively eliminates the potential delay of high-risk orders and significantly reduces the overall probability of order overdue in the system.

[0052] 3. The total delivery slack of the workshop is defined as the sum of the delivery slack of all orders. With the goal of balancing the delivery margin of each workshop, a better solution to the distributed workshop scheduling on-time completion problem is quickly obtained, which greatly reduces the algorithm difficulty and computational cost. It is suitable for complex heterogeneous manufacturing scenarios. It adopts intelligent optimization algorithm for global search and local search based on production system simulation. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the order allocation method of the present invention;

[0054] Figure 2 This is a schematic diagram of the process of order fulfillment at factory 1 according to the present invention;

[0055] Figure 3 This is a schematic diagram of the process of order processing at factory 2 according to the present invention;

[0056] Figure 4 This is a schematic diagram of the process of order processing at factory 3 according to the present invention;

[0057] Figure 5 This is a schematic diagram showing the delivery dates for each order in this invention;

[0058] Figure 6 This is a schematic diagram of the encoding of the genetic algorithm of the present invention;

[0059] Figure 7 This is a schematic diagram of the overall model of factory 1 according to the present invention;

[0060] Figure 8 This is a schematic diagram of the process modeling of order A in factory 1 in the simulation model of this invention;

[0061] Figure 9 Gantt charts of each factory in the simulation model after global optimization of order allocation in this invention;

[0062] Figure 10 This is a Gantt chart of each factory in the simulation model after the local adjustment of order allocation in this invention. Detailed Implementation

[0063] The present invention will be further described below with reference to the accompanying drawings and embodiments, which illustrate the above and other technical features and advantages of the present invention. However, the following embodiments are merely preferred embodiments of the present invention and are not exhaustive.

[0064] Example 1:

[0065] like Figure 1 As shown, this invention provides a simulation-based distributed heterogeneous machining workshop order allocation method and system. The order allocation method specifically includes the following steps:

[0066] S1. Construct discrete event simulation models based on the actual conditions of each heterogeneous machining workshop;

[0067] S2, the order allocation scheme is globally optimized by a genetic algorithm based on single-layer coding. The genetic algorithm takes maximizing the minimum delivery slack value among all workshops as the objective function to determine the optimal workshop assignment for each order;

[0068] S3. According to the order allocation plan, each order is input into the discrete event simulation model of the workshop. In each workshop, the production sequence of the orders is determined based on the delivery date. The production execution process is simulated. After running the simulation, the delivery date of each order is obtained, and the simulation verification is realized.

[0069] S4. Based on the delivery dates of each order in the simulation results, determine whether the on-time completion rate of the orders meets the requirements. If the requirements are met, output the final order allocation plan and the order production sequence of each workshop. Otherwise, perform local optimization, execute the order reallocation operation, identify the order with the longest delay, and migrate the overdue order to the target workshop with processing capacity based on the status output by the simulation model of each workshop.

[0070] S5, Update the order allocation plan based on the order migration results;

[0071] S6. Repeat steps S3-S5 until the maximum number of iterations is met or the on-time completion rate of the order meets the requirements.

[0072] In this embodiment, step S1 involves modeling the production systems of various heterogeneous distributed machining workshops using the Bowei Mozi production line simulation software to construct a discrete event simulation model. This specifically includes the following steps:

[0073] S11, create entity model objects based on the equipment layout and resource list of each distributed machining workshop, including processing equipment, buffer, tooling resources, worker pool, warehouse, loaders, assembly station, disassembly station, conveyor belt, AGV, RGV and overhead crane; then define the logistics scheduling module and worker scheduling module; set actual characteristics, including machine failure characteristics;

[0074] S12, Import process information, including process time, process sequence, resource requirements, worker requirements, transfer tools, batch size and product model for each product;

[0075] S13, edit the script in the discrete event simulation model to ensure that each product runs according to the actual process route and process time;

[0076] S14, collect the order information to be processed, simulate and verify the module, and put it into production in the following order: the order with the closer delivery date is put into production first. If the delivery dates of two orders are the same, the order with the smaller delivery date slack is put into production first.

[0077] The initial state of the discrete event simulation model is set according to the real-time status of the workshop. The entire process from the start of production to the completion of all orders is simulated. Key performance indicators are recorded during the process, including the actual completion time of each order, equipment utilization rate, key bottleneck processes, Gantt chart, work-in-process analysis and part throughput time. The "on-time completion rate of orders" is evaluated, which is the proportion of orders completed on time to the total number of orders.

[0078] In this embodiment, in step S2, the chromosomes are numbered using a single-layer coding method, with each chromosome representing an order allocation scheme to the workshop. The gene value of the _th gene locus represents the _th For each order, the assigned workshop number ensures that all gene values ​​are within the range of valid workshop numbers.

[0079] When it exists One pending order and When there are candidate workshops, the set of orders to be assigned Candidate workshop set Each order With the customer's required delivery date. Each order In the workshop The fastest completion time required is Based on this, the steps for global optimization are as follows:

[0080] S21, Chromosome Encoding and Population Initialization: An order allocation scheme is represented using integer encoding; the chromosome length is equal to the total number of orders. , No. Gene value of each gene locus , indicating an order Assigned to the workshop ,Right now ;

[0081] Randomly generated Each initial chromosome constitutes the initial population, with each gene value... exist Randomly select a value within a range;

[0082] S22, Calculate the workshop delivery slack and fitness values: For each chromosome in the population, according to the chromosome encoding, Each order was divided into Each workshop was assigned to a specific workshop. The order set is Delivery slack for each order The difference between the delivery date and the fastest completion time, i.e. Then the workshop The delivery slack is Then the fitness function of the genetic algorithm is: ;

[0083] S23, Genetic Operations: Selection uses tournament selection, crossover uses single-point crossover, and mutation uses a low-probability mutation of each gene locus, randomly replacing gene values ​​with... The other integer in the sequence represents the order being randomly reassigned to another workshop;

[0084] S24, Population Update and Iteration: Adopting an elite retention strategy, repeat steps S21 to S23. After reaching the maximum number of iterations, terminate the iteration and output the order allocation scheme, which is the scheme with the lowest overall delay risk obtained by global search.

[0085] In this embodiment, during step S4, when performing local optimization and order reallocation:

[0086] A1. Based on the simulation results, identify overdue orders and their key bottleneck processes in the original workshop, thereby determining the key factors that lead to low on-time completion rates. For example, if multiple orders are delayed due to excessive load in a certain workshop, then that workshop is identified as a bottleneck node; if an order cannot be completed on time in the currently assigned workshop due to excessive queuing time, then that order is an object to be optimized.

[0087] A2. Based on the above judgment results, the initial order allocation plan is partially optimized. During the optimization, the remaining effective capacity of the candidate workshop, the current delivery slack, and the compatibility with the process route of the order are evaluated based on the real-time status output by the simulation model of each workshop. Then, several orders with delays are selected and the delayed orders are moved from the original allocation workshop to the workshop with the highest comprehensive evaluation score of the three indicators among other workshops. During the optimization process, process compatibility and equipment idle time are considered to ensure that the adjusted plan is feasible, so as to maximize the overall on-time completion rate while minimizing the adjustment cost.

[0088] A3. Input the optimized order allocation scheme into the simulation software again for verification. If the simulation results show that the on-time completion rate has reached the preset target, then the scheme is issued to each distributed workshop as the final allocation scheme for execution. Otherwise, continue with A1 and repeat A1 to A3 until the maximum number of iterations is met or the on-time completion rate of the orders meets the requirements.

[0089] Through a closed-loop optimization process of "simulation evaluation - problem diagnosis - local adjustment - re-verification", the high computational cost and execution risk caused by global reallocation are effectively avoided, achieving efficient and dynamic optimization of distributed workshop order allocation, and significantly improving the responsiveness and reliability of the production system.

[0090] In this embodiment, the order allocation system includes a simulation modeling module, an algorithm optimization module, a simulation verification module, an order migration module, and a scheduling coordination engine;

[0091] The simulation modeling module is used to build and maintain discrete event simulation models for multiple heterogeneous machining workshops;

[0092] The algorithm optimization module uses a genetic algorithm to randomly generate and optimize the initial order allocation scheme.

[0093] The simulation verification module is used to simulate the production execution process and obtain the delivery dates for each order;

[0094] The order migration module is used to perform order reassignment operations;

[0095] The scheduling and coordination engine is used to update orders and input simulation models to obtain simulation results.

[0096] Example 2:

[0097] like Figure 2-10 As shown, 10 orders need to be processed in 3 distributed heterogeneous workshops. The processing technology for each order in each workshop is as follows: Figure 2 , Figure 3 and Figure 4 As shown, the delivery time for each order is as follows: Figure 5 As shown;

[0098] D1, obtains the initial order allocation scheme through a genetic algorithm: one set of codes is as follows Figure 6 As shown, after decoding, the order for processing in workshop 1 is {A,F,G,J}, and the delivery slack is:

[0099] ;

[0100] The order for processing in Workshop 2 is {B,E,H}, and the delivery slack is:

[0101] ;

[0102] The order for processing in workshop 3 is {C,D,I}, and the delivery slack is:

[0103] ;

[0104] The fitness function of the genetic algorithm, where the fitness value of this encoding is 135;

[0105] With the population size set to 300, crossover probability at 0.2, and mutation probability at 0.01, the globally optimized order allocation scheme is: Workshop 1 processes {A,C,E,G}, Workshop 2 processes {F,H,I}, and Workshop 3 processes {B,D,J}, with a fitness value of 147.

[0106] D2. Simulation models of workshops 1, 2, and 3 were constructed using the Bowei Mozi production line simulation software. The overall model of workshop 1 is as follows: Figure 7 As shown, it includes layout modeling, data modeling, and logical modeling. The process information modeling for order A is as follows: Figure 8 As shown;

[0107] After running the simulation model, the completion time of workshop 1 was found to be 107 for A, 124 for C, 79 for E, and 70 for G.

[0108] In workshop 2, the completion time of F is 114, the completion time of H is 77, and the completion time of I is 75.

[0109] In workshop 3, the completion time for B is 70, the completion time for D is 75, and the completion time for J is 143. Figure 9 As shown, the delivery date for order J is 140 days, indicating a delay.

[0110] D3. Adjustments were made to the processing workshop for overdue order J. Simulation results showed that workshop 2 had a lower processing load. An attempt was made to transfer order J to workshop 2. The results showed that, with other order delivery dates remaining unchanged, the delivery date of order J was brought forward to 140 days. Figure 10 As shown, the requirements are met, and the iteration ends.

[0111] The above description is merely a preferred embodiment of the present invention and is illustrative rather than restrictive. Those skilled in the art will understand that many changes, modifications, and even equivalents can be made within the spirit and scope defined by the claims of the present invention, all of which will fall within the protection scope of the present invention.

Claims

1. A method for order allocation in a distributed heterogeneous machine shop based on simulation, characterized in that, The order allocation method specifically comprises the following steps: S1. Constructing a discrete event simulation model according to the actual situation of each isomerization machine shop; S2. Globally optimizing the order allocation scheme by a single-layer coding-based genetic algorithm, the genetic algorithm taking the maximum of the minimum tardiness relaxation value in all shops as an objective function to determine the optimal shop attribution of each order; S3. According to the order allocation scheme, inputting each order into the discrete event simulation model of the shop, determining the production order of each order according to the delivery date in each shop, simulating the production execution process, and obtaining the delivery date of each order after running the simulation to realize simulation verification; S4. Judging whether the on-time completion rate of the order meets the requirements through the delivery date of each order in the simulation result, outputting the final order allocation scheme if the requirements are met, otherwise performing local optimization and executing order re-allocation operation, identifying the longest overdue order, and migrating the overdue order to the target shop with processing capacity based on the state output by the simulation model of each shop; S5. Updating the order allocation scheme according to the order migration result; S6. Repeating steps S3-S5 until the maximum iteration number is reached or the on-time completion rate of the order meets the requirements.

2. The simulated-based distributed heterogeneous machine shop order assignment method of claim 1, wherein, The step S1 specifically comprises the following steps: S11. Creating entity model objects according to the equipment layout and resource list of each distributed machine shop, including processing equipment, buffer zone, tooling resource, worker pool, warehouse, loader, assembly station, disassembly station, conveyor belt, AGV, RGV and overhead crane, defining the logistics scheduling module and the worker scheduling module, and setting the actual characteristics; S12. Importing process information, including process time, process sequence, resource demand, worker demand, transfer tool, batch size and product model of each product; S13. Editing scripts in the discrete event simulation model to ensure that each product is simulated and run according to the actual process route and process time; S14. Collecting order information to be processed, and simulating and verifying the module according to the following order: the order with the nearest delivery date is produced first, and if the delivery dates of two orders are the same, the order with the smallest tardiness relaxation is produced first; Setting the initial state of the discrete event simulation model according to the real-time state of the shop.

3. The simulated-based distributed heterogeneous machine shop order assignment method of claim 1, wherein, In the step S2, the single-layer coding method is: Each chromosome represents an assignment scheme of orders to workshops, the chromosome length is equal to the total number of orders, the gene value at the th gene position represents the workshop number assigned to the th order, all gene values are in the range of valid workshop numbers.

4. The simulated-based distributed heterogeneous machine shop order assignment method of claim 1, wherein, In the step S2, the objective function is specifically: For each order assigned to a shop, the difference between the due date of the order and the earliest completion time at the shop is the due date slack , in particular: ； wherein, the delivery date of the order the delivery date, denotes the fastest finishing time required when only order n is processed in the current shop. the total due date slack for each shop is the sum of the due date slack for all orders assigned to the shop That is: ； wherein, represents the set of all orders assigned to the i-th workshop. Initialize algorithm parameters and order information, minimum factory due date slack is fitness value i.e. ； The objective function is to maximize the minimum total tardiness relaxation in all shops, that is: ； Wherein, m represents the total number of shops, which is used to improve the overall completion robustness of the system and avoid local shop over-tension.

5. The simulated-based distributed heterogeneous machine shop order assignment method of claim 1, wherein, In the step S3, the simulation verification comprises the following steps: S31. Simulating the scheduling execution according to the order process route and resource constraints in the simulation model, and outputting the actual completion time of each order; S32. If there is an order with a completion time later than its delivery date, it is determined as an overdue order, triggering the subsequent order re-allocation mechanism; S33. If the overall on-time completion rate reaches the preset threshold or reaches the upper limit of the number of order re-allocation, it is confirmed that the current order allocation scheme is feasible.

6. The simulated-based distributed heterogeneous machine shop order assignment method of claim 1, wherein, In the step S4, the order re-allocation operation comprises the following steps: S41. Identifying the overdue order and its key bottleneck process in the original shop; S42, based on the real-time state output by each workshop simulation model, evaluate the three indexes of the candidate workshop, including the remaining effective capacity, the current delivery time slack, and the compatibility with the order process route; S43, select the workshop with the highest score in the comprehensive evaluation of the three indexes as the target workshop, execute the order migration, and update the production plan of the relevant workshops; S44, re-run the simulation to verify the adjusted overall scheme.

7. A system for the simulation-based distributed heterogeneous machine shop order allocation method according to any one of claims 1 to 6, characterized in that The order allocation system comprises a simulation modeling module, an algorithm optimization module, a simulation verification module, an order migration module, and a scheduling coordination engine; The simulation modeling module is used to build and maintain discrete event simulation models of multiple heterogeneous machine workshops; The algorithm optimization module randomly generates an initial order allocation scheme and implements optimization through a genetic algorithm; The simulation verification module is used to simulate the production execution process and obtain the delivery time of each order; The order migration module is used to execute the order re-allocation operation; The scheduling coordination engine is used to update the order and input the simulation model to obtain the simulation result.

Images