Order process merging workshop scheduling method and system based on job level constraint

Abstract

The invention relates to an order process merging workshop scheduling method and system based on operation level constraints. The method comprises the steps of obtaining production data; establishing a mixed integer programming scheduling model taking the minimization of the maximum completion time and the minimization of the assembly time as targets; order combination constraints, process combination constraints and operation level constraints are integrated in the mixed integer programming scheduling model; and an improved grey wolf optimization algorithm is established to solve the mixed integer programming scheduling model, an optimal scheduling scheme is output, the improved grey wolf optimization algorithm combines a double-layer coding scheme of three constraints, a mixed initialization population strategy and an improved hunting mechanism, and a variable neighborhood search and simulated annealing dynamic acceptance mechanism is integrated, so that the optimal scheduling scheme of the hybrid integer programming scheduling model is obtained. Global exploration and local development capabilities are enhanced, so that an optimal scheduling scheme is obtained; and finally outputting the dispatching Gantt chart and issuing and executing. The scheduling problem caused by a complex assembly relation and too long preparation time in multi-variety small-batch production is effectively solved, and the production period is remarkably shortened.

Description

A Shop Floor Scheduling Method and System Based on Job Hierarchy Constraints for Merging Order Processes Technical Field

[0001] This invention relates to the field of intelligent manufacturing and production scheduling technology, specifically to a method and system for merging order processes and scheduling workshops based on job level constraints. Background Technology

[0002] With the advent of Industry 4.0 and smart manufacturing, the manufacturing production model is rapidly shifting from traditional mass production to multi-variety, small-batch production. In this model, the workshop simultaneously handles a large number of small-scale orders with varying delivery dates, and products are typically complex assemblies composed of multiple levels of intermediate parts. This complexity leads to two core challenges: First, due to the flexibility and diversity of products, the preparation time for tool changes and material handling during the processing of workpieces accounts for a significantly increased proportion of the total production cycle, while the actual processing time is relatively small. Second, due to the assembly dependencies between manufactured parts, operations are subject to hierarchical priority constraints, significantly increasing the complexity of workshop scheduling.

[0003] While significant progress has been made in flexible job shop scheduling research, most studies focus on designing efficient solution algorithms or simple batching strategies based on product attributes. Research on integrated scheduling problems involving the triple constraints of "job hierarchy constraints," "order merging," and "process merging" remains insufficient. Traditional scheduling methods typically assume independent processes and that the processing sequence between parts is not constrained by assembly relationships, or consider only a single merging strategy, making it difficult to meet the actual needs of multi-variety, small-batch manufacturing workshops. This results in existing methods producing low-quality initial solutions, easily getting trapped in local optima, and failing to obtain ideal optimized scheduling schemes when facing such complex problems, thus failing to effectively shorten production cycles and reduce setup time.

[0004] Therefore, there is an urgent need for a flexible job shop scheduling method that can effectively handle job level constraints and collaboratively optimize order merging and process merging in order to improve the overall efficiency of the shop in the multi-variety, small-batch production mode. Summary of the Invention

[0005] The purpose of this invention is to provide a workshop scheduling method and system for merging order processes based on job level constraints. By establishing a scheduling model that integrates multiple constraints and designing an efficient solution algorithm, it achieves a significant reduction in production preparation time through an intelligent merging strategy while satisfying complex assembly relationships, thereby optimizing the maximum completion time and assembly time.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: an order process merging workshop scheduling method based on job hierarchy constraints, used for scheduling in flexible workshops, comprising the following steps:

[0007] S1. Obtain production data;

[0008] S2. Establish a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time and minimizing the assembly time; the mixed-integer programming scheduling model integrates:

[0009] Order consolidation constraint: Consolidates multiple orders with delivery date differences within a preset threshold into the same batch for processing;

[0010] Process merging constraint: Multiple processes that are processed consecutively on the same machine will be merged and executed to form a continuous processing block;

[0011] Operation hierarchy constraint: For workpieces with assembly relationships, all operations of the sub-component workpiece are processed after all operations of the main component workpiece are completed;

[0012] S3. Establish an improved gray wolf optimization algorithm to solve the mixed-integer programming scheduling model and output the optimal scheduling scheme.

[0013] S4. Based on the optimal scheduling scheme, generate a workshop production scheduling Gantt chart and issue it for execution.

[0014] Preferably, in S1, the production data includes an order set, a set of workpieces contained in each order, a set of processes contained in each workpiece, a set of machines, a set of assembly units, the processing time and preparation time of each process on an optional machine, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces.

[0015] Preferably, the order consolidation constraint further includes:

[0016] Each order can only be assigned to one batch;

[0017] The order quantity for each batch shall not exceed the batch size;

[0018] Each workpiece must be processed before its latest delivery time;

[0019] The start time of the first operation for all workpieces in an order must be later than the start time of the batch;

[0020] The batch size of each batch is dynamically calculated based on the machine's maximum processing capacity within a unit of time and the processing time of each process.

[0021] Preferably, the process merging constraint further includes:

[0022] Each process is carried out only on permitted machines;

[0023] The completion time for each process includes its processing time and preparation time;

[0024] The start time of the subsequent process shall not be earlier than the completion time of the preceding process;

[0025] Only one process can be processed on the same machine within the same time slot.

[0026] More preferably, the objective functions for order merging constraints and process merging constraints are:

[0027] ;

[0028] ;

[0029] This refers to the overall completion time. This is the first optimization objective. It is time. It is a collection of time. and It is a 0-1 variable. It is an order Delay penalty / delay time, It is a batch index. It is the collection of all processing batches. It is the collection of all production orders.

[0030] Preferably, the job level constraint further includes: job level constraint: for workpieces with assembly relationships, all operations of the sub-component workpiece are processed after all operations of the main component workpiece are completed.

[0031] Each order must have one and only one assembly unit constraint;

[0032] Only one order can be processed in the same assembly unit within the same time slot.

[0033] If a process needs to be processed after assembly, then the start time of its subsequent process must be later than the assembly completion time.

[0034] Assembly completion time equals start time plus assembly time.

[0035] More preferably, the objective function for the job hierarchy constraint is:

[0036] ;

[0037] ;

[0038] in, This is the maximum value for all assembly completion times. It is the second optimization objective; It is an order At the assembly station Assembly completion time, It is a collection of assembly stations. It is a batch index. It is the collection of all processing batches.

[0039] Preferably, the improved gray wolf optimization algorithm includes:

[0040] A two-layer encoding method combining process codes and machine codes is adopted, in which the generation and decoding process of process codes naturally embeds the rules of order merging, process merging and job hierarchy constraints;

[0041] Establish a hybrid initialization strategy: use a combination of heuristic rules and completely random generation to generate the initial population.

[0042] Preferably, the hybrid initialization strategy includes:

[0043] 50% of the individuals in the population are generated using heuristic rules, which are a combination of delivery date priority and shortest processing time priority.

[0044] The other 50% of individuals in the population are generated in a completely random manner, only needing to satisfy the process route sequence constraint;

[0045] Apply a random perturbation with a preset probability to the individuals generated by the heuristic rules.

[0046] Preferably, the improved gray wolf optimization algorithm further includes constructing an improved hunting mechanism to update the individual gray wolf positions, the improved hunting mechanism including:

[0047] Crossover operation: Randomly select two of the current three leader wolves (Alpha, Beta, Delta) as parents, and perform segmented crossover to generate offspring individuals;

[0048] Mutation operation: The mutation probability is dynamically adjusted according to the iteration progress, and the operation of swapping, insertion or reversal is performed on the process code of the offspring individual.

[0049] Preferably, the improved gray wolf optimization algorithm further includes establishing a variable neighborhood search and simulated annealing mechanism, specifically:

[0050] For the three leader wolves Alpha, Beta, and Delta, different neighborhood structures are used for local search. These neighborhood structures include: swapping the positions of two random steps, inserting a random step into a new random position, and reversing a random subsequence of consecutive steps. For new solutions generated through variable neighborhood search, a simulated annealing mechanism is used to dynamically determine whether to accept them. The probability P of accepting inferior solutions is calculated using the following formula:

[0051] ;

[0052] ;

[0053] in: , t is the current iteration number.

[0054] This invention also discloses a job level constraint-based order process merging workshop scheduling system, which is used for a job level constraint-based order process merging workshop scheduling method. The system includes:

[0055] The data acquisition module produces data packages. The production data includes order sets, the set of workpieces contained in each order, the set of operations contained in each workpiece, machine sets, assembly unit sets, the processing time and preparation time of each operation on the available machines, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces.

[0056] The model building module constructs a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time and minimizing the assembly time, and includes order merging, process merging, and job hierarchy constraints.

[0057] The optimization solution module is configured with an improved gray wolf optimization algorithm, and the improved gray wolf optimization algorithm is called to solve the mixed integer programming scheduling model and obtain the optimal scheduling scheme.

[0058] The scheduling output module outputs the optimal scheduling scheme in the form of a Gantt chart and sends it to the workshop production execution system.

[0059] Beneficial effects: This invention integrates job hierarchy constraints, order merging, and process merging within the constructed mixed integer programming scheduling model, and combines this with an improved gray wolf optimization algorithm; it achieves a significant reduction in production preparation time through intelligent merging strategies while satisfying complex assembly relationships, thereby optimizing maximum completion time and assembly time; it is more in line with the actual production scenarios of complex assembly workshops with multiple varieties and small batches.

[0060] In this invention, the improved Grey Wolf optimization algorithm effectively solves the problems of large solution space and easy getting trapped in local optima under complex constraints by using double-layer encoding, hybrid initialization, improved hunting mechanism, and deep integration of variable neighborhood search and simulated annealing. Experiments show that the algorithm significantly outperforms the traditional NSGA-II algorithm in terms of solution quality, convergence speed, and stability.

[0061] In addition, by intelligently merging orders and processes, this invention effectively reduces preparation time (such as tool changing and material handling) during workpiece processing, ensures smooth assembly by satisfying operational level constraints, and ultimately significantly shortens the maximum completion time and assembly time, thereby improving equipment utilization and overall workshop production efficiency. Attached Figure Description

[0062] Figure 1 is a flowchart of the flexible workshop scheduling method provided in an embodiment of the present invention.

[0063] Figure 2 is a flowchart of the improved Grey Wolf Optimization Algorithm (IGWO) in an embodiment of the present invention.

[0064] Figure 3 is an example diagram of the process code in an embodiment of the present invention.

[0065] Figure 4 is a Gantt chart of unconstrained process scheduling according to an embodiment of the present invention.

[0066] Figure 5 is a Gantt chart of scheduling after process constraints according to an embodiment of the present invention.

[0067] Figure 6 is a schematic diagram of the segmented cross-operation of the improved hunting mechanism in an embodiment of the present invention.

[0068] Figure 7 is a Gantt chart of machine tasks for a small-scale IGWO instance according to an embodiment of the present invention.

[0069] Figure 8 is a schematic diagram of the algorithm iteration process for small-scale instances according to an embodiment of the present invention.

[0070] Figure 9 is a Gantt chart of machine tasks in a medium-scale example of the IGWO embodiment of the present invention.

[0071] Figure 10 is a schematic diagram of the algorithm iteration process for medium-scale instances according to an embodiment of the present invention.

[0072] Figure 11 is a Gantt chart of the IGWO machine task in a large-scale example of an embodiment of the present invention.

[0073] Figure 12 is a schematic diagram of the large-scale instance algorithm iteration process according to an embodiment of the present invention. Detailed Implementation

[0074] To make the objectives and advantages of this invention clearer, the invention will be specifically described below with reference to embodiments. It should be understood that the following text is merely used to describe one or more specific embodiments of the invention and does not strictly limit the scope of protection specifically claimed by the invention.

[0075] The core of this invention lies in solving the scheduling challenges arising from complex assembly relationships (work hierarchy constraints) and lengthy production preparation times in multi-variety, small-batch manufacturing workshops. This invention optimizes production efficiency through innovations in both modeling and algorithms.

[0076] Example: Referring to Figure 1, the order process merging workshop scheduling method based on job hierarchy constraints is used for scheduling in flexible job workshops, and includes the following steps:

[0077] Step 1: Obtain production data;

[0078] Production data includes order sets, the set of workpieces contained in each order, the set of operations contained in each workpiece, machine sets, assembly unit sets, processing time and setup time for each operation on the available machines, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces.

[0079] Step 2: Establish a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time (Makespan) and minimizing the assembly time; the mixed-integer programming scheduling model integrates:

[0080] Order consolidation constraint: Sets delivery date differences within a preset threshold. Multiple orders are merged into the same batch for processing to reduce the total number of processing batches and equipment preparation times. The objective function is:

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] ;

[0089] in, It is an integer variable; it represents the order index, indicating different production orders. It is an integer variable representing the batch index, which represents the merged processing batch; This is the first optimization objective. It is a comprehensive indicator variable. It is a unit of time; T is a set of time. and , It is a 0-1 variable; , It is an order Delay penalty / delay time; It is the collection of all processing batches; It is the collection of all production orders; It is the maximum allowable latency difference threshold; BatchSizeb is the batch size. The actual capacity; It is an order Resource consumption / occupancy; MaxBatchSize is the maximum capacity limit of a batch; It is an order The number of subtasks included; It is an order A set of processes; It is an order process resource consumption, It is equipment For the process Processing efficiency / resource adaptability coefficient It is an order process Processing time, It is an order process Preparation time, It is equipment The maximum resource carrying capacity, It is a collection of equipment. It is an order process The start time of the first sub-process. It is the maximum relaxation constant.

[0090] Process merging constraint: Multiple processes that can be processed consecutively on the same machine are merged into a continuous processing block to reduce equipment changeover frequency and setup time. The objective function is:

[0091] ;

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] It is a 0-1 variable. It is an order process In the equipment On the processing priority, It is an order The total number of processes, where M is the set of equipment. It is a 0-1 parameter. It is an order process Completion time on device k It is an order process In the equipment The start time on It is an order process Processing time, It is an order process Preparation time, It is an order A set of processes.

[0099] Operational hierarchy constraints: For workpieces with assembly relationships, it is strictly stipulated that all operations of the sub-component workpiece can only begin after all operations of the main component workpiece are completed, ensuring the logical feasibility of the production process. The objective function is:

[0100] ;

[0101] ;

[0102] ;

[0103] ;

[0104] ;

[0105] ;

[0106] ;

[0107] in, The second optimization objective is ζ, which is a comprehensive indicator for the assembly process. It is an order At the assembly station Assembly completion time, It is a collection of assembly stations. It is a 0-1 variable. It is an order At the assembly station Assembly start time. It is an order The set of pre-processing steps, It is an order process In the equipment The processing completion time, where PostAssemblyij is a 0-1 parameter. It is an order Subsequent processes In the equipment The start time, It is an order At the assembly station Assembly time.

[0108] Step 3: Establish an improved Grey Wolf Optimization Algorithm to solve the mixed-integer programming scheduling model and output the optimal scheduling scheme. Refer to Figure 2, which shows the flowchart of the improved Grey Wolf Optimization Algorithm (IGWO). The improved Grey Wolf Optimization Algorithm includes:

[0109] (1) Two-layer coding scheme: The two-layer coding method combining process code and machine code is adopted, in which the generation and decoding process of process code naturally embeds the rules of order merging, process merging and operation hierarchy constraints;

[0110] The two-layer coding scheme is as follows:

[0111] The first segment is the process code, the length of which is equal to the total number of processes for all workpieces. The number of each workpiece is repeated according to the number of its processes. The sequence of this segment indicates the processing order of the processes.

[0112] The second segment is the machine code, which is the same length as the process code. The value in each position represents the processing machine number selected for the corresponding process.

[0113] Furthermore, rules for order merging, process merging, and job hierarchy constraints are embedded in the process of generating and decoding the process code.

[0114] (2) Establish a hybrid initialization strategy: The initial population is generated by combining heuristic rules (delivery date priority and shortest processing time priority) with completely random generation, taking into account both the quality and diversity of solutions;

[0115] The hybrid initialization strategy is as follows:

[0116] 50% of the individuals in the population are generated using heuristic rules, which are a combination of delivery date priority and shortest processing time priority.

[0117] The other 50% of individuals in the population are generated in a completely random manner, only needing to satisfy the process route sequence constraint;

[0118] Apply random perturbations with preset probabilities to individuals generated by heuristic rules to enhance population diversity;

[0119] (3) Improved hunting mechanism: Employ segmented crossover and adaptive probabilistic mutation operations to update the individual gray wolf positions to adapt to discrete scheduling problems; the improved hunting mechanism includes:

[0120] Crossover operation: Randomly select two of the current three leader wolves (Alpha, Beta, Delta) as parents, and perform segmented crossover to generate offspring individuals;

[0121] Mutation operation: The mutation probability is dynamically adjusted according to the iteration progress, and the operation of swapping, insertion or reversal is performed on the process code of the offspring individual.

[0122] (4) Establishing a variable neighborhood search and simulated annealing mechanism: Integrating a variable neighborhood search mechanism to apply various neighborhood operations to the leader wolf; and embedding a simulated annealing dynamic acceptance mechanism to accept inferior solutions with a certain probability, thereby enhancing the algorithm's ability to escape local optima; specifically:

[0123] For the three leader wolves Alpha, Beta, and Delta, different neighborhood structures are used for local search. These neighborhood structures include: swapping the positions of two random steps, inserting a random step into a new random position, and reversing a random subsequence of consecutive steps. For new solutions generated through variable neighborhood search, a simulated annealing mechanism is used to dynamically determine whether to accept them. The probability P of accepting inferior solutions is calculated using the following formula:

[0124] ;

[0125] ;

[0126] in: , t is the current iteration number.

[0127] This invention also discloses a job level constraint-based order process merging workshop scheduling system, which is used for a job level constraint-based order process merging workshop scheduling method. The system includes:

[0128] The data acquisition module produces data packages. The production data includes order sets, the set of workpieces contained in each order, the set of operations contained in each workpiece, machine sets, assembly unit sets, the processing time and preparation time of each operation on the available machines, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces.

[0129] The model building module constructs a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time and minimizing the assembly time, and includes order merging, process merging, and job hierarchy constraints.

[0130] The optimization solution module is configured with an improved gray wolf optimization algorithm, and the improved gray wolf optimization algorithm is called to solve the mixed integer programming scheduling model and obtain the optimal scheduling scheme.

[0131] The scheduling output module outputs the optimal scheduling scheme in the form of a Gantt chart and sends it to the workshop production execution system.

[0132] In a specific case study: taking the workshop of a power plant valve manufacturing company as the application background:

[0133] Step 1: Obtain production data for a specific period from the enterprise's Manufacturing Execution System. The data includes:

[0134] Order information: 13 (small-scale instance) production orders, each order includes the workpieces and quantities, and the latest delivery date.

[0135] Process information: There are 50 processes in total, including the processing time, preparation time, and optional processing machines for each process.

[0136] Resource information: List of available processing machines, assembly unit information.

[0137] Constraint information: The assembly dependencies between workpieces are determined through the bill of materials and process routes, forming a task hierarchy constraint tree. For example, order 2 (sub-component) can only begin processing after all operations of order 1 (main component) are completed.

[0138] Step 2: The objective function is to minimize the maximum completion time. And minimize the maximum assembly completion time ζ, and apply order merging constraints, process merging constraints, and job hierarchy constraints;

[0139] Step 3: Solve using the improved Grey Wolf optimization algorithm:

[0140] Encoding and Decoding: A two-layer encoding method as shown in Figure 3 is adopted. The process code not only represents the processing sequence, but its structure itself implicitly satisfies the three major constraints through specific rules (such as merging the process codes of orders that need to be merged adjacently and fixing the process codes with hierarchical constraints in order). During decoding, in addition to translating the process sequence and machine allocation, the start and end times of each process also need to be calculated, and a Gantt chart is finally generated.

[0141] Referring to Figure 3, Figure 3(a) is an example of a work code without constraints; Figure 3(b) is an example of a work code with order merging constraints applied; Figure 3(c) is an example of a work code with process merging constraints applied; Figure 3(d) is an example of a work code with job level constraints applied; and Figure 3(e) is a complete coding example that finally integrates the three constraints.

[0142] Decoding the process codes and machine codes in Figure 3(a) that lack process constraints yields the Gantt chart shown in Figure 4. Comparing Figures 4 and 5, it can be seen that constraining processes based on order merging, process merging, and the hierarchical constraints between workpieces reduces preparation time between processes and shortens the workshop production cycle.

[0143] Population initialization: A hybrid strategy is employed; 50% of individuals are generated using heuristic rules: orders are first sorted by delivery date, then processes within the same order are arranged according to their sequence, and the machine with the shortest processing time is selected for each process. The other 50% of individuals are generated completely randomly. This approach ensures a balance between a high-quality starting point and population diversity.

[0144] Improved Hunting Mechanism: The traditional continuous position update of GWOs is not suitable for scheduling problems. This invention adopts the segmented crossover shown in Figure 6: two wolves are randomly selected from Alpha, Beta, and Delta wolves, randomly segmented, and their gene fragments are exchanged to generate offspring. The mutation operation dynamically adjusts the probability, performing operations such as swapping and insertion on the action code.

[0145] Variable Neighborhood Search and Simulated Annealing: In each iteration, the Alpha, Beta, and Delta wolves are searched using the N1 (exchange), N2 (insertion), and N3 (reversal) neighborhood structures shown in Figure 5, respectively. New solutions are accepted if they are better than the current solution; otherwise, they are rejected according to the formula... The calculated probability is acceptable, where the temperature T decays with iteration. This effectively avoids premature convergence.

[0146] Archive and Elite Preservation: Maintain a global historical optimal solution archive and update the Alpha, Beta, and Delta three-headed leader wolves in each generation to ensure that the optimal solution is not lost and is used to guide the population.

[0147] Experimental Results and Analysis:

[0148] To verify the effectiveness of this invention, three sets of test cases were conducted: small, medium, and large scale. The Gantt charts of the machine tasks obtained by the two algorithms in their fifth iteration are shown in Figure 7. The Gantt charts of the medium-scale machine tasks with optimal completion times for both algorithms are shown in Figure 9. The Gantt charts of the large-scale machine tasks with optimal completion times for both algorithms are shown in Figure 11. As shown in Figures 8, 10, and 12 and the experimental data table, the IGWO algorithm converges faster and is more stable. This demonstrates the superiority of the proposed model and algorithm in solving such complex scheduling problems.

[0149] The embodiments of the present invention have been described in detail above with reference to the examples. However, the present invention is not limited to the above embodiments. For those skilled in the art, after learning the contents described in the present invention, several equivalent changes and substitutions can be made without departing from the principle of the present invention. These equivalent changes and substitutions should also be considered to fall within the protection scope of the present invention.

Claims

1. A job-level constraint-based order-process merging workshop scheduling method for flexible workshop scheduling, characterized by: The process includes the following steps: S1, acquiring production data; S2, establishing a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time and minimizing the assembly time; the mixed-integer programming scheduling model integrates: order merging constraint: merging multiple orders with delivery date differences within a preset threshold into the same batch for processing; process merging constraint: merging multiple processes that are processed continuously on the same machine to form a continuous processing block; job hierarchy constraint: for workpieces with assembly relationships, all processes of the sub-component workpiece are processed after all processes of the main component workpiece are completed; S3, establishing an improved gray wolf optimization algorithm to solve the mixed-integer programming scheduling model and outputting the optimal scheduling scheme; S4, generating a workshop production scheduling Gantt chart based on the optimal scheduling scheme and issuing it for execution.

2. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 1, characterized in that: In S1, the production data includes an order set, a set of workpieces contained in each order, a set of processes contained in each workpiece, a set of machines, a set of assembly units, the processing time and preparation time of each process on an optional machine, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces.

3. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 1, characterized in that: The order consolidation constraints also include: each order can only be assigned to one batch; the order quantity of each batch cannot exceed the batch size; each workpiece must be completed before its latest delivery time; the start time of the first operation of all workpieces in an order must be later than the start time of the batch; and the batch size of each batch is dynamically calculated based on the machine's maximum processing capacity and the processing time of the operation within a time unit.

4. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 3, characterized in that: The process merging constraints also include: each process is only processed on an authorized machine; the completion time of each process includes its processing time and preparation time; the start time of the subsequent process is not earlier than the completion time of the preceding process; and only one process can be processed on the same machine in the same time slot.

5. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 1, characterized in that: The job level constraints also include: job level constraints: for workpieces with assembly relationships, all operations of the sub-component workpiece are processed after all operations of the main component workpiece are completed. Each order must be assigned to one and only one assembly unit constraint; the same assembly unit can only process one order in the same time slot; if an operation needs to be processed after assembly, the start time of its subsequent operation must be later than the assembly completion time; the assembly completion time is equal to the start time plus the assembly time.

6. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 1, characterized in that: The improved Grey Wolf optimization algorithm includes: adopting a two-layer encoding method that combines process code and machine code, wherein the generation and decoding process of process code naturally embeds the rules of order merging, process merging and job hierarchy constraints; and establishing a hybrid initialization strategy: using a combination of heuristic rules and completely random generation to generate the initial population.

7. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 6, characterized in that: The hybrid initialization strategy includes: 50% of the individuals in the population are generated using heuristic rules, which are a combination of delivery date priority and shortest processing time priority; the other 50% of the individuals in the population are generated using a completely random method, only needing to satisfy the process route sequence constraint; and a preset probability random perturbation is applied to the individuals generated by the heuristic rules.

8. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 1, characterized in that: The improved gray wolf optimization algorithm also includes constructing an improved hunting mechanism to update the position of individual gray wolves. The improved hunting mechanism includes: crossover operation: randomly selecting two from the current three leader wolves (Alpha, Beta, Delta) as parents and performing segmented crossover to generate offspring individuals; mutation operation: dynamically adjusting the mutation probability according to the iteration progress and performing swap, insertion, or reverse operations on the process codes of offspring individuals.

9. The order process merging workshop scheduling method based on job hierarchy constraints according to claim 8, characterized in that: The improved gray wolf optimization algorithm also includes establishing a variable neighborhood search and simulated annealing mechanism. Specifically, different neighborhood structures are used for local search on the three leader wolves (Alpha, Beta, and Delta). These neighborhood structures include: swapping the positions of two random steps, inserting a random step into a new random position, and reversing a random continuous step subsequence. For new solutions generated through the variable neighborhood search, a simulated annealing mechanism is used to dynamically determine whether to accept them. The probability P of accepting inferior solutions is calculated using the following formula: ； ;in: ， t is the current iteration number.

10. A job-level constraint-based order-process merging workshop scheduling system, used to implement the job-level constraint-based order-process merging workshop scheduling method according to any one of claims 1-9, characterized in that: The system includes: The data acquisition module produces data packages. Production data includes order sets, the set of workpieces in each order, the set of operations in each workpiece, machine sets, assembly unit sets, processing time and setup time for each operation on the available machines, the latest delivery date of the order, and job hierarchy constraints representing the assembly dependencies between workpieces. The model building module constructs a mixed-integer programming scheduling model with the objectives of minimizing the maximum completion time and minimizing the assembly time, and includes order merging, operation merging, and job hierarchy constraints. The optimization solution module is configured with an improved gray wolf optimization algorithm, and the improved gray wolf optimization algorithm is called to solve the mixed integer programming scheduling model and obtain the optimal scheduling scheme. The scheduling output module outputs the optimal scheduling scheme in the form of a Gantt chart and sends it to the workshop production execution system.

Images