Resource-constrained flexible job shop scheduling method and system integrated with advanced production organization

Abstract

The invention relates to a resource-constrained flexible job shop scheduling method and system integrated with an advanced production organization, and the method comprises the steps: obtaining production data, building a scheduling model through the production data, and setting a corresponding key constraint; the scheduling model comprises an RCFJSP-SST model for integrating separable preparation time, an RCFJSP-SSTOBP model for integrating process batch processing, an RCFJSP-SSTPST model for integrating parallel sequential transmission, and an RCFJSP-SSTOBPPST model; and modeling the scheduling model by adopting a constraint programming model, and solving by utilizing a constraint programming solver to obtain an optimized scheduling scheme. According to the method, through collaborative integration and optimization of three advanced production organization modes, the manufacturing period is remarkably shortened, the equipment and resource utilization rate is increased, and an efficient and feasible scheduling scheme is provided for a multi-variety small-batch production scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent manufacturing and production scheduling, and particularly relates to a resource-constrained flexible job shop scheduling method and system integrated with advanced production organization. BACKGROUND

[0002] Flexible Job Shop Scheduling Problem (FJSP) is one of the most complex and representative NP-hard problems in manufacturing systems. The core task is to reasonably arrange the processing sequence and resource allocation for multiple workpieces on multiple optional machines under the constraints of process sequence, machine capacity, etc., in order to optimize key performance indicators such as manufacturing cycle (Makespan).

[0003] Existing researches on FJSP mainly evolve in three directions: 1) developing more advanced solving algorithms (such as meta-heuristic algorithms, reinforcement learning); 2) considering more complex production environments (such as distributed workshops, production-logistics collaboration); 3) introducing more realistic constraints and objectives (such as multi-resource constraints). However, most of the existing researches are still based on traditional production organization methods, mainly including: (1) Non-Separable Setup Times (NSST): the preparation work must start after the previous process is completely finished, resulting in long idle time of equipment and unnecessary extension of manufacturing cycle.

[0004] (2) Job-Specific Production: strictly maintaining the mapping relationship between workpieces and processes, under the background of popularization of modular design, the same front-end process cannot be combined, resulting in frequent switching of auxiliary resources (such as molds, clamps), and loss of scale economy effect.

[0005] (3) Sequential Transfer (ST): the next process must wait for all tasks of the previous process to be completed before starting, and its strict serial characteristics result in long waiting time of work-in-process and incompact production flow.

[0006] While a few studies have attempted to introduce Separable Setup Times (SST) or batching, these have typically been treated as isolated constraints, failing to explore the synergistic potential among the three advanced production organization methods: SST, Operation-Batch Processing (OBP), and Parallel-Sequential Transfer (PST). In resource-constrained environments, how to model and efficiently solve the integrated scheduling problem of these three methods has become a theoretical bottleneck and practical challenge for improving modern manufacturing efficiency.

[0007] Therefore, there is an urgent need for a scheduling method and system that can integrate SST, OBP and PST to fully leverage their combined advantages of reducing idle time, reducing preparation times and improving process parallelism, thereby achieving significant optimization of core indicators such as manufacturing cycle time. Summary of the Invention

[0008] To address the problems existing in the prior art, the present invention aims to provide a resource-constrained flexible workshop scheduling method and system that integrates advanced production organization. It aims to significantly shorten the manufacturing cycle, improve equipment and resource utilization, and provide an efficient and feasible scheduling solution for multi-variety, small-batch production scenarios through the synergistic integration and optimization of three advanced production organization modes.

[0009] To achieve the above objectives, the present invention provides the following solution: To achieve the above objectives, the present invention also provides a resource-constrained flexible job shop scheduling method integrating advanced production organization, characterized by comprising: Acquire production data, construct a scheduling model using the production data, and set corresponding key constraints; the scheduling model includes: the RCFJSP-SST model integrating separable preparation time, the RCFJSP-SSTOBP model integrating process batch processing, the RCFJSP-SSTPST model integrating parallel sequential transmission, and the RCFJSP-SSTOBPPST model. The scheduling model is modeled using a constraint programming model and solved using a constraint programming solver to obtain an optimized scheduling scheme.

[0010] Optionally, constructing the RCFJSP-SST model includes: Based on the RCFJSP model of resource-constrained flexible workshops, a separable preparation time mechanism is introduced to coordinate the resource allocation between the processing and preparation phases, ensuring that preparation time does not block the processing flow.

[0011] Optionally, the key constraints of the RCFJSP-SST model include: Triggering conditions: Trigger settings when consecutive tasks require different auxiliary resources: ; in, In preparation for triggering the flag, The total number of machines. When the value is 1, it means that the i-th machine was selected for the h-th process of the j-th workpiece; otherwise, it is 0. When the value is 1, it means that the h-th process of the j-th workpiece and the h'-th process of the j'-th workpiece use the same mold; otherwise, it is zero. When the value is 1, it means that on the i-th machine, the h-th process of workpiece j is processed before the h'-th process of workpiece j'. For machine indexing, For process index, For workpiece index, Let J be the total number of processes for workpiece j. Preparation start time: Begin when both machine and auxiliary resources are available. ; ; in, In preparation for the start, Let j'h' be the time to complete the preceding operation j'h' of the h-th operation on machine i. A value of 1 indicates that on the i-th machine, the h-th operation of workpiece j is performed before the h'-th operation of workpiece j'. To indicate the rth mold, For the complete set of molds, For the quantity of the r-th mold, For the d-th mold; Preparation duration: determined by the auxiliary resource setting time. ; in, For completion time, For the mold changeover time of the r-th mold, When the value is 1, it indicates that mold r is used in the h-th process of workpiece j.

[0012] Optionally, constructing the RCFJSP-SSTOBP model includes: Based on the RCFJSP-SST model, a batch processing mechanism is introduced to form a multi-branch flow structure by reconstructing the process network, and binary parameters are used to define the branching relationships between tasks. ; ; ; in, For the task Summing over all subsequent tasks, For the task The summation of the preceding tasks equals zero; For the set of all tasks, For processing tasks, When it is 1, it indicates a task. Supply to the task '; For the task The start time, For the task The completion time, It is a sufficiently large positive integer.

[0013] Optionally, constructing the RCFJSP-SSTPST model includes: Based on the RCFJSP-SST model, a parallel sequential transmission mechanism is introduced to control the number and size of sub-batches by minimizing the transmission sub-batch size. ; ; ; in, The number of sub-batch, The minimum size of the transmission sub-batch, For the batch size of workpiece j, Let be the batch size of the b-th sub-batch of item j.

[0014] Optionally, the key constraints of the RCFJSP-SSTPST model include: Sub-batch partitioning: Calculate the number and size of sub-batches based on the minimum transmission sub-batch; Transmission synchronization: The subsequent process is triggered by the completion time of the sub-batch. ; ; in, Let h be the start time of the h-th process for workpiece j. Let b be the completion time of the b-th transfer sub-batch in the h-1th process of workpiece j. It is a sufficiently large positive integer. When the value is 1, it means that the processing time of the h-th process of workpiece j is greater than the processing time of the previous process; This is the sequence number of the transmitted sub-batch. Let h be the start time of the h-th process for workpiece j. Let h be the completion time of the (h-1)th operation for workpiece j. Let h be the processing time for the h-th operation of workpiece j. Let b be the batch size of the b-th transmission sub-batch of workpiece j.

[0015] Optionally, constructing the RCFJSP-SSTOBPPST model includes: Based on the RCFJSP-SSTOBP model, a PST mechanism is introduced to handle transmission synchronization in multi-branch networks using flow segmentation priority rules, based on SPOSO weights and sub-batch allocation parameters. ; ; in, For the task The start time, For the task Preceding tasks The completion time of the b'th sub-batch; This is used to compare the processing times of related tasks. A value of 1 indicates that the processing time of the preceding task is shorter, and a value of 0 indicates otherwise. A value of 1 indicates the previous task. The The sub-batch has a mapping relationship with the b-th sub-batch of the subsequent task O; For processing task indexing, , To process sub-batch indexes, Tasks The number of sub-batches, For the task Processing time, For the task The size of the sub-batch of the b-th sub-batch.

[0016] Optionally, the key constraints of the RCFJSP-SSTOBPPST model include: Sub-batch quantity calculation: Considering downstream task requirements: ; in, , For the task The number of sub-batches, For the task The parameter indicates whether there are subsequent tasks. If it is 1, it means there is a task. There is a subsequent task; otherwise, there is no subsequent task. For the task The total set, When it is 1, it indicates a task. Supply to the task ', For the task Processing batch size; Sub-batch size allocation: allocated according to priority: ; ; in, For the task The size of the sub-batch of the b-th sub-batch; For the task The parameter indicates whether there are subsequent tasks. If it is 1, it means there is a task. There is a subsequent task; otherwise, there is no subsequent task. For the task The size of the sub-batch of the 'b'th sub-batch; A value of 1 indicates the previous task. The The sub-batch has a mapping relationship with the b-th sub-batch of the subsequent task O.

[0017] Optionally, obtaining the optimized scheduling scheme includes: The scheduling model is modeled using a constrained programming model, and the optimal scheduling scheme is obtained by using a constrained programming solver with the objective function of minimizing the manufacturing cycle. The constrained programming model is obtained by modeling interval variables, sequence variables, and constraint functions related to resource occupancy and timing. The constraint functions include: resource occupancy constraints, machine sequence constraints, task allocation constraints, and timing constraints.

[0018] To achieve the above objectives, the present invention also provides a resource-constrained flexible job shop scheduling system integrating advanced production organization, comprising: The data input module is used to acquire production data; The model building and selection module is used to build a scheduling model using the production data and set corresponding key constraints; the scheduling model includes: the RCFJSP-SST model integrating separable preparation time, the RCFJSP-SSTOBP model integrating process batch processing, the RCFJSP-SSTPST model integrating parallel sequential transmission, and the RCFJSP-SSTOBPPST model. The solution engine module is used to model the scheduling model using a constraint programming model and solve it using a constraint programming solver to obtain an optimized scheduling scheme.

[0019] The beneficial effects of this invention are as follows: Significantly improved scheduling performance: This invention effectively reduces equipment idle time and work-in-process waiting time through the collaborative integration of SST, OBP, and PST. Experiments show that the optimal model (RCFJSP-SSTOBPPST) can shorten the manufacturing cycle by an average of 10% and reduce the algorithm convergence time by an average of 45%.

[0020] High solution quality and stability: The constraint programming method used in this invention can obtain solutions that are better than or equivalent to those of the hybrid metaheuristic algorithm (GA-VNS, WOA-VNS) and mixed integer programming in 90.7% of the test cases, demonstrating excellent solution quality and stability.

[0021] Highly practical: This invention originates from real industrial scenarios such as home appliance manufacturing. The various resource constraints and advanced production modes considered are widely representative. The scheduling scheme provided is easy to deploy and implement in the production system and has high engineering application value.

[0022] High flexibility: This invention provides four optional scheduling models, allowing enterprises to flexibly choose the most suitable model for phased implementation based on their own production process characteristics and management foundation, thus lowering the application threshold. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a resource-constrained flexible job shop scheduling method integrating advanced production organization according to an embodiment of the present invention; Figure 2 This is a schematic diagram comparing the separable preparation time and the non-separable preparation time in an embodiment of the present invention; (a) is the non-separable preparation time, showing that the preparation work must be performed sequentially after the processing; (b) is the separable preparation time, showing that the preparation work can be performed in parallel with the preceding processing. Figure 3 The diagram shows a comparison of the process paths before and after the implementation of the batch processing mechanism in this embodiment of the invention; (a) is the original path before implementation, showing frequent resource switching; (b) is the optimized path after implementation, showing that processes with the same resources are merged into consecutive batches. Figure 4This is a comparative diagram of sequential transmission and parallel sequential transmission in an embodiment of the present invention; (a) is sequential transmission, showing the strict serial relationship between processes; (b) is parallel sequential transmission, showing that parallel processing between processes is achieved through sub-batch segmentation. Figure 5 The following are performance comparison charts of four scheduling models in representative instances according to embodiments of the present invention: (a) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 3; (b) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 5; (c) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 8; (d) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 20; (e) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 30; (f) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 50; (g) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 80; (h) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 100; (i) Performance comparison chart of four scheduling models in representative instances when the number of workers is equal to 120. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] like Figure 1 As shown, this embodiment discloses a resource-constrained flexible job shop scheduling method integrating advanced production organization, including: acquiring production data, constructing a scheduling model using the production data, and setting corresponding key constraints; the scheduling model includes: an RCFJSP-SST model integrating separable preparation time, an RCFJSP-SSTOBP model integrating process batch processing, an RCFJSP-SSTPST model integrating parallel sequential transmission, and an RCFJSP-SSTOBPPST model; the scheduling model is modeled using a constraint programming model, and solved using a constraint programming solver to obtain an optimized scheduling scheme.

[0028] This embodiment discloses a resource-constrained flexible job shop scheduling method integrating advanced production organization, including: S1: Data Input and Problem Modeling: Receive production data and construct a scheduling model based on the production data. The scheduling model is at least one of the following four models: RCFJSP-SST model integrating separable preparation time, RCFJSP-SSTOBP model further integrating process batch processing, RCFJSP-SSTPST model further integrating parallel sequential transmission, and RCFJSP-SSTOBPPST model integrating all three at the same time. S2: Model Solving: The scheduling model constructed in step S1 is modeled using the constraint programming method, and solved using the constraint programming solver to obtain the optimized scheduling scheme; S3: Scheduling Scheme Output: Output the optimized scheduling scheme.

[0029] The separable preparation time mechanism means that the preparation work required for a process can begin when the machine is idle and the necessary auxiliary resources are available, and can be executed in parallel with the processing task. The process batch processing mechanism means that processes with the same processing requirements in different workpieces are identified and merged to form continuous processing batches. The parallel sequential transfer mechanism means that the processes of a workpiece are divided into multiple transfer sub-batches, and after the current process completes a sub-batch, it is transferred to the next process to begin processing. In the parallel sequential transfer mechanism, the number and scale of sub-batches are controlled by setting a minimum transfer sub-batch size.

[0030] The constraint programming method described in step S2 uses interval variables, sequence variables, and a resource occupancy function to model the scheduling model. The resource occupancy function is a pulse function, used to constrain the total usage of auxiliary resources to not exceed their capacity. For the integrated batch processing model, a distance matrix including setup time is combined with a noOverlap constraint to define the time interval between adjacent tasks on the machine. The constraint programming solver described in step S2 is CPLEX CPOptimizer.

[0031] This embodiment discloses a resource-constrained flexible job shop scheduling method integrating advanced production organization, including the following steps: S1: Data Input and Problem Modeling: Receive production data including workpiece set, process sequence, machine set, auxiliary resource set, batch size of each workpiece, unit processing time of each process on each machine, and setup time and quantity of each auxiliary resource; based on this data, construct at least one of the following four scheduling models: Model 1 (RCFJSP-SST): A basic scheduling model that integrates separable preparation time; Structural Composition: This model introduces a Separable Setup Time (SST) mechanism on the basis of the Resource Constrained Flexible Job Shop (RCFJSP). SST allows the preparation work of a process to begin independently when the machine is idle and the required auxiliary resources are available, and to be executed in parallel with the processing task, thereby reducing machine idle time.

[0032] Functionality: Through mathematical modeling, resource allocation between the processing and preparation phases is coordinated to ensure that preparation time does not block the processing flow.

[0033] Key constraints include: Preparation trigger condition: Trigger the setup when consecutive tasks require different auxiliary resources (Formula 1). Preparation start time: Start when both the machine and auxiliary resources are available (Formulas 2-3). Preparation duration: Determined by the auxiliary resource setup time Tr (Formula 4).

[0034] (1); (2); (3); (4); (5); (6); in, and The start and finish times are respectively the preparation start time and the preparation finish time. Formula (5) stipulates that the start time of the h-th process of workpiece j shall not be earlier than the end time of its preparation work. Formula (6) is the global optimization objective, which stipulates that the completion time of all processes of all workpiece j shall not be earlier than the maximum completion time. Prepare to trigger the flag.

[0035] Model 2 (RCFJSP-SSTOBP): Based on Model 1, it further integrates process batch processing; Structural composition: The RCFJSP-SST model introduces a process batching (OBP) mechanism to merge the same processes in different workpieces into continuous processing batches, reducing the frequency of auxiliary resource switching.

[0036] Functionality: OBP reconstructs the process network to form a multi-branch flow structure (caused by the delay strategy) and uses the binary parameter PREOO′ to define the branching relationship between tasks (Equation 7-9).

[0037] Key constraints include: Batch merging: Identical processes are merged into a single task O, with cumulative processing time. Network structure: The number of predecessors and successors of a task is measured using SPOSO and SPREO. SPOSO is the sum of all subsequent tasks for task O; a value of 0 indicates that task O has no downstream tasks, and its value reflects the number of its subsequent tasks. SPOSO is the sum of all preceding tasks for task O; its value can only be 0 or 1. A value of 0 indicates that task O has no preceding tasks, and a value of 1 indicates that task O has only one preceding task.

[0038] (7); (8); (9); Here, PREOO′=1 indicates that task O branches to task O′.

[0039] Model 3 (RCFJSP-SSTPST): Based on Model 1, it further integrates parallel sequential transmission; Structural composition: The RCFJSP-SST model introduces a parallel sequential transfer (PST) mechanism, which divides the process into multiple transfer sub-batches, allowing the subsequent process to be triggered immediately after the completion of the previous process sub-batch.

[0040] Functionality: The number and size of sub-batches are controlled by setting the minimum transfer sub-batch size MinB (Formula 10-12), and processing continuity is ensured.

[0041] Key constraints include: Sub-batch partitioning: Calculate the number of sub-batches based on MinB. and size Transmission synchronization: Time is completed via sub-batch. Trigger the start of the next process (Formula 13-14).

[0042] (10); (11); (12); (13); (14); Model 4 (RCFJSP-SSTOBPPST): A complete scheduling model that integrates separate preparation time, batch processing of processes, and parallel sequential transmission.

[0043] Structure: Based on the RCFJSP-SSTOBP model, the PST mechanism is introduced, which combines the resource parallelism of SST, the batch optimization of OBP, and the transmission parallelism of PST.

[0044] Functionality: Based on SPOSO weights and sub-batch allocation parameters, and using stream segmentation priority rules, this function implements the following functionality: Handling transmission synchronization in multi-branch networks (Equations 18-19).

[0045] Key constraints include: Sub-batch quantity calculation: Considering downstream task requirements (Formula 15).

[0046] Sub-batch size allocation: allocated according to priority (Formula 16-17).

[0047] (15); (16); (17); (18); (19); in, This indicates whether task O has a downstream task.

[0048] S2: Model Solving: The scheduling model selected in step S1 is modeled using the constraint programming method, and solved using a constraint programming solver to obtain the scheduling scheme that minimizes the manufacturing cycle. The constraint programming model is constructed using interval variables, sequence variables, and constraint functions related to resource usage and timing. The specific process is as follows: Modeling process: Variable definition: Interval variables: represent the execution interval of a process. For example, Zjh represents the processing interval of process Ojh. Optional attribute is associated with machine assignment.

[0049] Sequence Variables: Represent the sequence of tasks on a machine. For example, SeqMi represents the sequence of all operations on machine i. The type attribute is associated with auxiliary resources.

[0050] Integer and binary variables: used to represent resource allocation, sequence order, etc.

[0051] Constraint functions: Resource occupancy constraints: The pulse function is used to manage auxiliary resource occupancy. For example, pulse(Zjh, 1) means that the resource intensity occupied by process Zjh is 1 (Formula 20).

[0052] Machine sequence constraints: The noOverlap function is used to ensure that tasks do not overlap on the same machine. For the OBP model, the time setting is defined in conjunction with the distance matrix DMOO′ (Formula 21).

[0053] Task assignment constraint: Use the alternative function to ensure that each task is assigned to only one available machine (Equation 22).

[0054] Timing constraints: Use functions such as endBeforeStart and startOf to define the sequence relationship between processes (Formulas 23 and 24).

[0055] Objective function: Minimize the manufacturing cycle Cmax, which is the maximum value of the completion time of all processes (Formula 25).

[0056] Solution process: Solver configuration: CPLEX CP Optimizer is used as the constraint programming solver, and the solution time limit (600 seconds for small-scale instances and 3600 seconds for large-scale instances) and iteration failure limit (1015 times) are set.

[0057] Solution steps: Model compilation: Convert the CP model into a solver executable format.

[0058] Search strategy: Based on branch and bound and constraint propagation techniques, prioritize high-impact variables (such as machine sequence and resource allocation).

[0059] Convergence criterion: The solution is terminated when the optimal solution (GAP=0) is reached or the time limit is reached.

[0060] Performance optimization: For large-scale instances, a heuristic initial solution and parallel search strategy are used to accelerate convergence.

[0061] (20); (twenty one); (twenty two); (twenty three); (twenty four); (25); S3: Scheduling Scheme Output: Output the optimal or near-optimal scheduling scheme, which includes the machines allocated to each process, the auxiliary resources used, the start time, the completion time, and the overall manufacturing cycle.

[0062] Solution formation mechanism: The scheduling solution is directly derived from the variable assignments after solving the constraint programming model. Machine allocation: The presenceOf attribute of the interval variable ZMijh determines whether process Ojh is allocated to machine i.

[0063] Auxiliary resource allocation: determined by the type attribute (associated auxiliary resource index) of the sequence variable SeqMi and the resource occupancy function pulse.

[0064] Time arrangement: The start time Sjh and end time Cjh of the process are determined by the startOf and endOf attributes of the interval variable.

[0065] Manufacturing cycle: The overall completion time is given by the objective function variable Cmax.

[0066] The scheduling plan outputs the following structured data: machine index and auxiliary resource index allocated to each process; start time and finish time of each process.

[0067] Inter-process transfer sub-batch division and transfer time points (applicable to PST model).

[0068] Statistics on overall manufacturing cycle Cmax and resource utilization rate.

[0069] Relationship with the model: The scheduling scheme is a feasible solution obtained from the model, satisfying all mathematical constraints (such as resource capacity, sequence order, transmission synchronization, etc.). Through the optimization search of the constraint programming solver, the scheme minimizes the manufacturing cycle while ensuring the feasibility and efficiency of production resources.

[0070] Furthermore, the separable preparation time mechanism means that the preparation work required for the process can begin when the machine is idle and the required auxiliary resources are available, without waiting for the previous process to be completed, thereby realizing the parallel execution of preparation work and processing work.

[0071] Furthermore, the process batch processing mechanism refers to identifying and merging identical processes with the same processing requirements in different workpieces to form a continuous processing batch, thereby reducing the switching frequency of auxiliary resources.

[0072] Furthermore, the parallel sequential transmission mechanism refers to dividing a single process of a workpiece into multiple transmission sub-batches; once the current process completes the processing of a sub-batch, it can be immediately transferred to the next process to begin processing, without waiting for all sub-batches of the process to be completed, thereby realizing parallel operation between processes.

[0073] Furthermore, in the parallel sequential transmission mechanism, a minimum transmission sub-batch size is set to control the number and scale of sub-batches in order to balance scheduling performance and logistics complexity.

[0074] Furthermore, the constraint programming solver is a CPLEX CP Optimizer.

[0075] This embodiment also discloses a resource-constrained flexible job shop scheduling system integrating advanced production organization, including: a data input module for acquiring production data; a model building and selection module for constructing a scheduling model using the production data and setting corresponding key constraints; the scheduling model includes: an RCFJSP-SST model integrating separable setup time, an RCFJSP-SSTOBP model integrating batch processing of processes, an RCFJSP-SSTPST model integrating parallel sequential transmission, and an RCFJSP-SSTOBPPST model; and a solution engine module for modeling the scheduling model using a constraint programming model and solving it using a constraint programming solver to obtain an optimized scheduling scheme.

[0076] Specifically, this embodiment also discloses a resource-constrained flexible job shop scheduling system integrating advanced production organization, including: Data input module: used to receive and store production data; Model building and selection module: used to select and build the corresponding constraint programming model from the four scheduling models based on user configuration or production characteristics; Solver Engine Module: Integrates a constraint planning solver for solving the constructed model; Results Output and Visualization Module: Used to output and display the final scheduling scheme.

[0077] Figure 2 This is a schematic diagram comparing the separable preparation time and the non-separable preparation time in an embodiment of the present invention; Figure 2 (a) For non-separate preparation time, the preparation work must be serial after the processing. It can only start after all the preceding work is completed. Figure 2 (b) For separable preparation time, the preparation work can be performed in parallel with the preceding processing. It does not need to wait for the preceding task to be completed; the preparation work can start as long as machine and mold resources are available. Obviously, Figure 2 (b) Separable preparation is more efficient and helps to shorten the production cycle; Figure 3 This is a comparison diagram of the process paths before and after the implementation of the batch processing mechanism in this embodiment of the invention; Figure 3 (a) shows the original path before implementation, which shows frequent resource switching. When different processes use the same mold, the scheduling scheme will cause the mold to switch back and forth. Figure 3 (b) To show the optimized path after implementation, the same resource processes are merged into consecutive batches, which can significantly reduce mold changeover, especially for large molds, where the changeover time can be as long as several hours. Figure 4 This is a schematic diagram comparing sequential transmission and parallel sequential transmission in an embodiment of the present invention; Figure 4 (a) is a sequential transmission, which shows the strict serial relationship between processes. It requires that the processing of the next process can only begin after all the previous processes are completed. Figure 4 (b) Parallel sequential transmission demonstrates how to achieve parallel processing between processes through sub-batch segmentation. The next process can begin processing as soon as the previous process is partially completed and resources are available. Figure 5 This is a performance comparison chart of four scheduling models in representative instances according to embodiments of the present invention; CP1~CP4 in the chart correspond to RCFJSP-SST, RCFJSP-SSTOBP, RCFJSP-SSTPST, and RCFJSP-SSTOBPPST, respectively. Figure 5 (a)-(i) correspond to work quantities of 3, 5, 8, 20, 30, 50, 80, 100, and 120, respectively. Figure 5 (a)-(c) are defined as small-scale examples. Figure 5 (d)-(f) are small-scale examples in the definition, and (g)(h)(i) are large-scale examples. Figure 5 (a)-(c) show that CP3 performs best in small-scale examples, but as the scale increases, CP3 gradually loses its advantage, while CP2 and CP4, which integrate OBP technology, gradually show their advantages, such as... Figure 5 (d)-(i). Ultimately, CP4, which integrates both OBP and PST technologies, has the best overall performance, reducing the target value by 10% and the solution time by 45% compared to the original scheduling mode CP1.

[0078] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method of integrating advanced production organization with resource-constrained flexible job shop scheduling, characterized by, The method comprises the following steps: acquiring production data, constructing a scheduling model by using the production data, and setting corresponding key constraints; the scheduling model comprises an RCFJSP-SST model integrating separable setup time, an RCFJSP-SSTOBP model integrating process batch processing, an RCFJSP-SSTPST model integrating parallel sequential transfer, and an RCFJSP-SSTOBPPST model; a constraint programming model is used to model the scheduling model, and a constraint programming solver is used to solve the scheduling model to obtain an optimized scheduling scheme.

2. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The RCFJSP-SST model is constructed by: introducing a separable setup time mechanism based on an RCFJSP model of a resource-constrained flexible job shop, which is used to coordinate the resource allocation of the processing phase and the preparation phase, and ensure that the setup time does not block the processing flow.

3. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The key constraints of the RCFJSP-SST model are set by: a preparation trigger condition: triggering the setting when consecutive tasks require different auxiliary resources; ； wherein, is a preparation trigger flag, is a total number of machines, is 1 when the i-th machine is selected for the h-th process of the j-th workpiece, otherwise 0; is 1 when the h-th process of the j-th workpiece uses the same mold as the h'-th process of the j'-th workpiece, otherwise 0; is 1 when the h-th process of the j-th workpiece precedes the h'-th process of the j'-th workpiece on the i-th machine; is a machine index, is a process index, is a workpiece index, is a total number of processes of the j-th workpiece; a preparation start time: starting when the machine and the auxiliary resource are available; ； ； wherein, for preparing to start, for indicating the completion time of the immediately preceding process j'h' of the hth process of the workpiece j on the machine i, for indicating that the hth process of the workpiece j on the ith machine is prior to the h'th process of the workpiece j' when the value is 1, for indicating the rth mold, for the total set of molds, for the number of the rth mold, for the dth set of molds; a preparation duration: determined by the auxiliary resource setting time; ； wherein, is the changeover time of the rth die, is the changeover time of the rth die, is the hth process of the workpiece j using the rth die.

4. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The RCFJSP-SSTOBP model is constructed by: introducing a process batch processing mechanism based on the RCFJSP-SST model, which is used to form a multi-branch flow structure by reconstructing the process network, and define the branch relationship between tasks using a binary parameter: ； ； ； in, For the task Summing over all subsequent tasks, For the task The summation of the preceding tasks is zero; otherwise, it is zero. For the set of all tasks, For processing tasks, When it is 1, it indicates a task. Supply to the task '; For the task The start time, For the task The completion time, It is a sufficiently large positive integer.

5. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The RCFJSP-SSTPST model is constructed by: introducing a parallel sequential transfer mechanism based on the RCFJSP-SST model, which is used to control the number and size of sub-batches by using the minimum transfer sub-batch size: ； ； ； wherein, is the number of sub-lots, is the size of the smallest transmission sub-lot, is the lot size of workpiece j, is the lot size of the bth sub-lot of workpiece j.

6. The resource-constrained flexible job shop scheduling method integrated with advanced production organization of claim 5, wherein, The key constraints of the RCFJSP-SSTPST model are set by: sub-batch division: calculating the number and size of sub-batches according to the minimum transfer sub-batch size; transport synchronization: triggering the next process by the sub-batch completion time: ； ； wherein, is the start time of the hth operation of the workpiece j, is the completion time of the bth transport sub-lot of the (h-1)th operation of the workpiece j, is a sufficiently large positive integer, is 1 if the processing time of the hth operation of the workpiece j is greater than the processing time of the previous operation; is the serial number of the transport sub-lot, is the start time of the hth operation of the workpiece j, is the completion time of the (h-1)th operation of the workpiece j, is the processing time of the hth operation of the workpiece j, is the lot size of the bth transport sub-lot of the workpiece j.

7. The resource-constrained flexible job shop scheduling method integrated with advanced production organization of claim 1, wherein, The RCFJSP-SSTOBPPST model is constructed by: introducing a PST mechanism based on the RCFJSP-SSTOBP model, which is used to process the transfer synchronization in the multi-branch network based on the SPOSO weight and the sub-batch allocation parameter by using the flow splitting priority rule: ； ； wherein, is the start time of the task , is the finish time of the task , is the finish time of the b'th sub-batch of the task , is 1 if the processing time of the preceding task is shorter than the processing time of the following task, otherwise 0; is 1 if the b'th sub-batch of the task , has a mapping relationship with the b'th sub-batch of the task O; is the processing task index, , is the processing sub-batch index, is the number of sub-batches of the task , is the processing time of the task , is the sub-batch size of the b'th sub-batch of the task .

8. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The key constraints of the RCFJSP-SSTOBPPST model are set by: sub-batch number calculation: considering the downstream task demand; ； wherein, , is the number of sub-lots of the task , is a parameter indicating whether there is a subsequent task for the task , if it is 1, it indicates that there is a subsequent task for the task , otherwise, there is no subsequent task; is the total set of tasks , is 1, it indicates that the task is supplied to the task , is the processing batch of the task ; sub-batch size allocation: according to the priority allocation: ； ； in, For the task The size of the sub-batch of the b-th sub-batch; For the task The parameter indicates whether there are subsequent tasks. If it is 1, it means there is a task. There is a subsequent task; otherwise, there is no subsequent task. For the task The size of the sub-batch of the 'b'th sub-batch; A value of 1 indicates the previous task. The The sub-batch has a mapping relationship with the b-th sub-batch of the subsequent task O.

9. The resource-constrained flexible job shop scheduling method integrating advanced production organization of claim 1, wherein, The optimized scheduling scheme is obtained by: using a constraint programming model to model the scheduling model, and using a constraint programming solver to solve the scheduling model with the manufacturing cycle as the objective function, to obtain an optimized scheduling scheme; the constraint programming model is modeled by using interval variables, sequence variables, and constraint functions related to resource occupation and timing; wherein the constraint functions comprise resource occupation constraints, machine sequence constraints, task allocation constraints, and timing constraints.

10. An integrated advanced production organization's resource-constrained flexible job shop scheduling system implemented according to the method of any one of claims 1-9, characterized in that, The method comprises the following steps: a data input module for acquiring production data; a model construction and selection module for constructing a scheduling model by using the production data, and setting corresponding key constraints; The scheduling model comprises: an RCFJSP-SST model integrating separable preparation time, an RCFJSP-SSTOBP model integrating process batch processing, an RCFJSP-SSTPST model integrating parallel sequential transmission, and an RCFJSP-SSTOBPPST model. A solving engine module is configured to model the scheduling model by using a constraint programming model, and solve the scheduling model by using a constraint programming solver to obtain an optimized scheduling scheme.

Images