Cabin part two-stage mixed job shop scheduling method considering beat difference

Abstract

The invention provides a cabin part two-stage mixed job shop scheduling method considering beat difference, and aims to improve the efficiency and accuracy of productivity evaluation. The method comprises the following steps: acquiring workpiece information and resource information of cabin part production and manufacturing; obtaining a production constraint and a target function according to the production characteristics of the two-stage hybrid job shop, and constructing a two-stage hybrid job shop scheduling model; designing an improved genetic algorithm to solve the model; and carrying out production resource scheduling according to an optimization result. According to the method, by considering the beat difference of cabin part production, a two-stage hybrid job shop scheduling method which is more suitable for the complex production task requirements of modern manufacturing enterprises is provided, the enterprises can quickly and effectively obtain a better scheduling scheme, and the workshop production efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of production workshop scheduling technology, and in particular to a two-stage mixed operation workshop scheduling method for cabin parts that takes into account cycle time differences. Background Technology

[0002] As core load-bearing structures in high-end equipment such as aerospace vehicles, the manufacturing quality and efficiency of cabin components directly affect the overall equipment's performance and development cycle. These components typically feature complex configurations, high material removal rates, and are produced in small batches with diverse varieties. Their manufacturing process generally employs a two-stage hybrid production model combining "assembly line roughing" and "parallel machine finishing." Under this model, the inherent differences in processing rhythm between processes can easily create dynamic bottlenecks, leading to problems such as equipment waiting for materials, work-in-process inventory buildup, and extended production cycles. Therefore, researching workshop scheduling methods that consider these rhythm differences in a two-stage hybrid operation is of significant theoretical importance and outstanding engineering application value for improving the balance of complex product manufacturing processes, shortening delivery cycles, and enhancing enterprises' market responsiveness.

[0003] For the scheduling problem of two-stage hybrid workshops, production processing often considers single machines or single production lines. However, in real-world production environments, multiple production lines and multiple machines are more common, which better reflects the coordination between production lines and parallel machines. Currently, most two-stage scheduling focuses on distributed and assembly workshop scenarios, and few existing studies consider combining permutation workshops with discrete parallel workshops. This is mainly because permutation workshops require all processes on the same production line to follow a consistent workpiece processing sequence, while in discrete parallel workshops, workpieces can flexibly choose processing paths on different machines. The two differ fundamentally in their constraint properties and scheduling rules, making it difficult to mathematically express an integrated model. Therefore, there is an urgent need for a scheduling method that can deeply integrate the characteristics of permutation workshops and discrete parallel workshops to support integrated decision-making for multi-line, multi-machine collaboration. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention is designed for two-stage mixed operation workshop production scenarios and constructs a scheduling model that considers the differences in production cycle time (i.e., the differences in processing time of different machines). It aims to achieve precise matching of production cycle time and optimization of the overall process, providing key technical support for cabin parts manufacturing enterprises to achieve lean production and efficient scheduling.

[0005] To achieve the above-mentioned technical objectives, the present invention provides the following technical solution: A two-stage hybrid operation workshop scheduling method for cabin parts that considers cycle time differences, specifically including: A two-stage hybrid workshop for the production of cabin parts is constructed, and workpiece information and resource information for cabin parts production are collected; the two-stage hybrid workshop is a hybrid production system consisting of a first-stage flow workshop and a second-stage parallel workshop. Based on the collected workpiece and resource information, as well as the production characteristics of the two-stage mixed operation workshop, a two-stage mixed operation workshop scheduling model is established, and the objective function and production constraints of the two-stage mixed operation workshop scheduling model are determined. An improved genetic algorithm was designed to solve the two-stage mixed-operation workshop scheduling model and obtain the optimal production scheduling scheme. Based on the optimal production scheduling scheme, the workpiece-assembly line allocation result, workpiece-parallel machine allocation result, workpiece processing sequence and processing time plan are obtained, which guides the two-stage production resource scheduling of cabin parts and completes the production task.

[0006] Furthermore, the workpiece information includes the number of parts produced and the processing steps for each part; the resource information includes the number of production lines in the first stage, the number of machines on the production lines, and the processing time of each machine on the production lines, the number of parallel machines in the second stage, and the processing time of the workpieces on each parallel machine.

[0007] Furthermore, the production characteristics of the two-stage hybrid operation workshop specifically include: The first phase is the replacement assembly line workshop, including... There are 10 production lines, each with its own... Each machine represents the steps required for workpiece processing. In each process, the workpieces on each machine on the same production line are processed in the same order, and each workpiece can only be processed on one production line at a time and the production line cannot be changed. The second phase is the discrete parallel workshop, including A parallel machine is available, allowing for flexible selection of the machine for processing workpieces. In production, each workpiece is required to complete its production task by sequentially passing through two stages of workshops.

[0008] Furthermore, the objective function is to minimize the maximum completion time of the production task, where the maximum completion time of the production task is the maximum value of the completion times of each parallel machine in the second stage.

[0009] Furthermore, the production constraints include workpiece-assembly line allocation uniqueness constraints, assembly line process sequence constraints, two-stage connection constraints, and parallel machine non-overlapping processing constraints; The workpiece-production line allocation uniqueness constraint is as follows: In the first stage, each workpiece can only be assigned to one position on one production line, and each position on each production line can only be occupied by one workpiece; in the global sequence of workpiece processing in the first stage, each workpiece at each position is assigned to one and only one production line. The specific constraints on the production line process sequence are as follows: in the production line Above, position The completion time of each workpiece on each machine shall not be less than the processing time of that machine; position The workpiece in the machine The completion time is no earlier than when it is installed on the machine. Completion time, location of the workpiece In the machine The completion time of the workpiece is no earlier than that of the workpiece preceding it in the machine. Completion time; the location The first step in the global sequence of workpiece machining in the first stage One location; The two-stage connection constraint is as follows: in the second stage, each workpiece is assigned to one and only one parallel machine; the start time of the second stage is not earlier than the completion time of the first stage. The non-overlapping processing constraint of the parallel machine is specifically defined as follows: for two workpieces assigned to the same parallel machine, the completion time of the second stage of the workpiece being processed is no later than the start time of the second stage of the other workpiece.

[0010] Furthermore, the improved genetic algorithm described above is used to solve the two-stage mixed-operation shop scheduling model to obtain the optimal production scheduling scheme, specifically including: Initialize input information, which includes a set of workpieces. Assembly line Assembly line process set Parallel machine set Processing time of each step on the assembly line Processing time of the workpiece on the parallel machine ; Construct a chromosome with a three-segment integer coding structure, and set the total length of each chromosome to the number of workpieces. Three times that of the previous three, with each encoded segment having a length of 1.5 times. The first segment of the code is a global sequential code, where the code position indicates the global order of workpiece processing in the assembly line, and the integer at each code position represents the workpiece number at the current position; the second segment of the code is a workpiece-assembly line allocation code, where the code position corresponds to the workpiece number, and the integer at each code position represents the assembly line number assigned to each workpiece; the third segment of the code is a workpiece-parallel machine allocation code, where the code position corresponds to the workpiece number, and the integer at each code position represents the parallel machine number assigned to the workpiece at the current position. Determine the parameters of the genetic algorithm, including population size, maximum number of iterations, crossover probability, and mutation probability; Determine the termination condition and fitness function of the genetic algorithm; the termination condition is: the current iteration number is greater than the maximum iteration number; calculate the objective function value of each individual based on the scheduling result, and use it as the input of the fitness function; The initial population was divided into two categories, with each category containing half of the total number of individuals in the population. The first category adopts a random strategy: each segment of the three-segment integer coding structure of the chromosome of each individual in the first category is generated randomly. The first segment of coding randomly generates the global order of workpiece processing in the first stage of the assembly line. The second segment of coding randomly assigns each workpiece to the assembly line according to the obtained global order. The third segment of coding then randomly assigns each workpiece to the parallel machine. The second approach uses the NEH+SPT strategy: First, according to the shortest time rule (SPT), each workpiece in the first-stage assembly line is assigned the assembly line with the shortest total processing time for each process, resulting in the second segment of the code; in the second-stage parallel workshop, each workpiece is assigned the parallel machine with the shortest processing time, resulting in the third segment of the code; finally, the total time for both stages is calculated, and the global order of workpiece processing in the first stage is generated using the NEH algorithm, resulting in the first segment of the code. Decode the individuals generated in each iteration to obtain a two-stage hybrid job shop scheduling scheme; carry out production operations according to the obtained scheduling scheme, and calculate the maximum completion time of the production task as the fitness of the current individual; Determine whether the population iteration has reached the preset maximum number of iterations. If it has not reached the maximum number of iterations, first save the best individual in the current iteration based on fitness, and then perform selection, crossover, and mutation operations on the population to obtain new individuals; calculate the fitness of the new individuals, compare it with the saved best individuals, and update the best individuals accordingly. If the maximum number of iterations is reached, the iteration stops, and the best individual that is finally saved is decoded to obtain the optimal scheduling scheme for the two-stage mixed operation workshop.

[0011] More specifically, the decoding of the individuals generated in each iteration to obtain the two-stage hybrid job shop scheduling scheme specifically includes: First, read the chromosomes from left to right. The encoding of the bits is used to obtain the global order of workpiece processing in the first stage; based on the chromosome number... Ranked first The integer at each encoded position assigns a pipeline to each workpiece, and according to the workpiece sequence in the global processing order of the first stage, each workpiece is placed to the corresponding pipeline for processing, resulting in a production list for each pipeline. and the processing time matrix of each production line The production list This includes the workpiece processing sequence on each production line, and the processing time matrix. Including workpieces On the assembly line No. Processing time on the machine ; Initialize pipeline information and place the workpiece On the assembly line No. Start time, completion time on the machine, workpiece The initial values ​​for the completion time of the first stage are all set to zero; then according to and The completion time of each workpiece on each machine on the production line is calculated recursively until the completion time of the first stage of each workpiece is obtained. Read chromosome number again To the The bit encoding determines the workpiece-parallel machine allocation result, and the production list of each parallel machine is obtained. and the processing time matrix of each parallel machine The production list The processing time matrix includes the workpieces assigned to each parallel machine. This includes the processing time of each workpiece on the parallel machine; Based on the completion time of each workpiece in the first stage, the actual processing sequence on each parallel machine is determined using a first-come, first-served rule. Calculate the start time and finish time of each workpiece on the parallel machine, and recursively calculate the finish time of each parallel machine based on the actual processing sequence. Take the maximum value of the finish times of the parallel machines as the maximum finish time of the current production task.

[0012] More specifically, the selection, crossover, and mutation operations performed on the population to obtain new individuals include: After preserving the best individuals from the previous generation, individuals are selected from the subpopulation through a binary tournament to enter the mating pool and produce the next generation; For the chromosomes of the next generation of individuals, crossover and mutation operations are performed, including: The crossover operation is as follows: sequential crossover is applied to the global sequential coding segment of the chromosome, and uniform crossover is applied to the workpiece-pipeline allocation coding segment and the workpiece-parallel machine allocation coding segment. The mutation operation involves randomly mutating all three segments of the chromosome's coding. New individuals are obtained through selection, crossover, and mutation operations.

[0013] Furthermore, this invention discloses an electronic device comprising a memory and a processor, wherein: Memory is used to store computer programs that can run on a processor; The processor is configured to, while running the computer program, execute a two-stage mixed operation workshop scheduling method for cabin parts that takes into account cycle time differences, as described above.

[0014] The present invention also discloses a computer-readable storage medium storing computer instructions for causing a processor to execute a two-stage mixed operation workshop scheduling method for cabin parts that takes into account cycle time differences, as described above.

[0015] Based on the above technical solution, the present invention has at least the following beneficial effects: The method proposed in this invention takes the two-stage collaboration of replacement flow shop and discrete parallel shop as the core of scheduling. By constructing a clear model structure and a reasonable scheduling process, it effectively reduces the waiting time between processes and the bottleneck problem of the flow line, realizes the precise matching of production cycle and the optimization of the overall processing flow. Based on the optimal scheduling scheme obtained by genetic algorithm, the cabin parts workshop can efficiently arrange production and clearly determine whether each workpiece can be delivered on time. The method proposed in this invention provides effective decision support for enterprises to deal with complex production tasks of multiple varieties and small batches. While improving the resource optimization and rapid response capability of workshop production, it provides key technical support for cabin parts manufacturing enterprises to achieve lean production and efficient scheduling. Attached Figure Description

[0016] Figure 1 This is a flowchart of a two-stage mixed operation workshop scheduling method for cabin parts that considers cycle time differences, as proposed in this invention. Figure 2 This is a flowchart of the genetic algorithm for solving the optimal production scheduling scheme in the method proposed in this invention; Figure 3 This is a schematic diagram of chromosome encoding for solving the optimal production scheduling scheme using the genetic algorithm in the method proposed in this invention. Detailed Implementation

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the following description is provided in conjunction with the accompanying drawings. Figure 1-3 The present invention will be further described in detail with reference to the embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0018] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0019] This invention proposes a flowchart for a two-stage hybrid operation workshop scheduling method for cabin parts that considers cycle time differences. By rationally allocating and sequencing workpieces on the assembly line and parallel machines, it not only effectively improves resource utilization efficiency and overall task scheduling efficiency, but also provides a practically feasible solution for workshop production. Please refer to [link / reference]. Figure 1 The method specifically includes the following steps: A two-stage hybrid workshop for the production of cabin parts is constructed, and workpiece information and resource information for cabin parts production are collected; the two-stage hybrid workshop is a hybrid production system consisting of a first-stage flow workshop and a second-stage parallel workshop. In a preferred embodiment, the workpiece information includes the number of parts produced and the processing steps for each part; the resource information includes the number of production lines in the first stage, the number of machines on the production lines, and the processing time of each machine on the production lines, the number of parallel machines in the second stage, and the processing time of the workpiece on each parallel machine.

[0020] In this embodiment, the workpiece information and resource information of the constructed two-stage hybrid operation workshop production and manufacturing instance are shown in Table 1 below: Table 1. Workpiece Information and Resource Information

[0021] As shown in the table above, in this embodiment, the production task is set to produce 30 workpieces to be processed. The first stage of the assembly line workshop includes 4 assembly lines, each with 3 machines, representing the 3 processes required for workpiece processing. The second stage of the parallel workshop includes 5 parallel machines with different processing efficiencies.

[0022] The processing time (production cycle time) of each workpiece varies on different production lines and parallel machines. Considering the large data dimensionality, in this embodiment, the processing time of each workpiece on the production line and parallel machine is simulated using random number generation methods within the range of 10 to 20 minutes and 15 to 20 minutes, respectively. In actual production, the data can be written into the database and retrieved. Detailed processing time information for each workpiece is shown in Table 1 above. For example, the processing time of workpiece 1 on production line 1 is [16, 13, 17], indicating that if this workpiece is processed on production line 1, it needs to go through 3 processes, with processing times of 16 minutes, 13 minutes, and 17 minutes for the first, third, and fourth processes, respectively. The processing time of workpiece 1 on the parallel machine is [18, 16, 17, 18, 19], indicating that the processing time of this workpiece on the first to fifth machines in the second stage is 18 minutes, 16 minutes, 17 minutes, 18 minutes, and 19 minutes, respectively.

[0023] Based on the workpiece and resource information collected above, as well as the production characteristics of the two-stage mixed operation workshop, a two-stage mixed operation workshop scheduling model is established, and the objective function and production constraints of the two-stage mixed operation workshop scheduling model are determined. In this embodiment, the production characteristics of the two-stage hybrid operation workshop specifically include: The first phase is the replacement assembly line workshop, including... There are 10 production lines, each with its own... Each machine represents the steps required for workpiece processing. In each process, the workpieces on each machine on the same production line are processed in the same order, and each workpiece can only be processed on one production line at a time and the production line cannot be changed. The second phase is the discrete parallel workshop, including A parallel machine is available, allowing for flexible selection of the machine for processing workpieces. In production, each workpiece is required to complete its production task by sequentially passing through two stages of workshops.

[0024] The objective function of the two-stage hybrid job shop scheduling model is set as minimizing the maximum completion time of the production task, where the maximum completion time of the production task is the maximum value of the completion times of all parallel machines in the second stage; the formula for the objective function is expressed as: ; in, This indicates the maximum completion time for a production task.

[0025] Considering the production characteristics and differences between the two stages, this embodiment sets production constraints such as workpiece-assembly line allocation uniqueness constraint, assembly line process sequence constraint, two-stage connection constraint, and parallel machine non-overlapping processing constraint. The uniqueness constraint for workpiece-to-production line assignment is based on practical considerations. It addresses the characteristics of replacement production lines in the aerospace industry, where the manufacturing of certain cabin parts involves fixed process paths and high specialization. This constraint facilitates practical implementation. Specifically, in the first stage, each workpiece can only be assigned to one position on one production line, and each position on each production line can only be occupied by one workpiece. In the global processing sequence of the workpieces in the first stage, each workpiece at each position is assigned to exactly one production line. The formula is as follows: ; in, For a collection of workpieces, Indicates a single workpiece index. This represents the total number of workpieces. This is the position index number of the workpiece in the global sequence during the first stage; For 0-1 decision variables, when the workpiece Position in the global order hour, ,otherwise ; is a 0-1 decision variable, when the workpiece is at a certain position. Assigned to the assembly line When, ,otherwise ; The assembly line process sequence constraints ensure that workpieces must pass through each process sequentially on the assembly line, strictly following the actual physical processing flow. Specifically, on the assembly line... Above, position The completion time of the workpiece on each machine Not less than the processing time of the machine ;Location The workpiece in the machine Completion time No earlier than its time in the machine Completion time The position of the workpiece In the machine The completion time of the workpiece is no earlier than that of the workpiece preceding it in the machine. Completion time The location The first step in the global sequence of workpiece machining in the first stage One position.

[0026] It's important to clarify that the global sequence in the first stage is designed only for the first-stage assembly line workshop. It primarily represents the order in which all workpieces enter the entire processing flow and are processed within the assembly line. For each assembly line, workpieces are sequentially assigned to that line from the global sequence, forming the processing sequence on that line. Once a workpiece is assigned to a particular assembly line, all processes are completed on that line, and the process sequence for all workpieces on the same line is consistent. In the second stage, the parallel workshop, the specific processing order of workpieces on each parallel machine is determined by a dynamic first-come, first-served rule, no longer constrained by the global sequence of the first stage.

[0027] In this embodiment, the process sequence constraint of the production line can be expressed by the following formula: Firstly, the constraints are on the assembly line. Above, position The completion time of the workpiece on the first machine The processing time must be no less than that of the first process. : ; Then, based on the process sequence, the position on the assembly line is determined sequentially. The workpiece in the first The completion time of the machine and its first The relationship between the completion time of the machines: ; and location The workpiece and the workpiece in front of it at the 1st The relationship between the completion time of the machines: ; In the above formula, A collection of assembly line processes; for The front position; It is a sufficiently large positive integer; For 0-1 decision variables, when position The workpieces are assigned to the assembly line When, ,otherwise ; Based on the uniqueness constraint of workpiece-to-production line allocation and the sequence constraint of production line operations, the completion time of the first stage can be linearized, as expressed by the formula: ; That is, according to the Big M method, if the workpiece Assigned to position and location Assigned to the assembly line Then the workpiece Completion time of the first phase It must be equal to its position on the assembly line. Completion time on the last machine This determines the completion time of the workpiece in the first stage, and the processing sequence of the second stage is then determined based on the completion time of the first stage.

[0028] In this embodiment, the second stage of production processing is based on two-stage connection constraints and parallel machine non-overlapping processing constraints. The two-stage connection constraints, from a mathematical model perspective, couple the two production stages—the assembly line workshop and the parallel workshop—into a whole to achieve global scheduling optimization across both stages. Specifically, in the second stage, each workpiece is assigned to one and only one parallel machine; the start time of the second stage is no earlier than the completion time of the first stage; the formula is expressed as: ; in, For 0-1 decision variables, when the workpiece Assigned to parallel machine When, ,otherwise ; This is the start time of the workpiece in the second stage; The non-overlapping processing constraint of the parallel machine ensures that the processing tasks on the same parallel machine conform to the actual production situation. Specifically, for two workpieces assigned to the same parallel machine, the completion time of the second stage of the workpiece being processed is no later than the start time of the second stage of the other workpiece; the formula is expressed as: ; in, For a set of parallel machines, For the workpiece set and Different workpieces; For 0-1 decision variables, when the workpiece and All allocated to parallel machines Above, and the workpiece In the workpiece During previous processing, ,otherwise ; , respectively workpiece , workpiece The completion time of the second stage, and in this embodiment, the completion time of the second stage satisfies the following formula: ; in, For workpiece Second-stage parallel machine The processing time.

[0029] The scheduling model for the two-stage hybrid workshop is now complete. This embodiment will first solve the scheduling model using traditional methods, such as the Gurobi solver. In this embodiment, Gurobi is used to verify the model's effectiveness. With small-scale data, the Gurobi solver can obtain an exact solution. However, as the data scale increases, the Gurobi solver's efficiency falls far short of actual production requirements. Therefore, this embodiment further designs a heuristic greedy algorithm to solve the scheduling model. The specific process of the heuristic greedy algorithm includes: (1) Traverse all workpieces to be processed and independently evaluate the processing performance of each workpiece on each production line. Specifically, calculate the workpiece... The total processing time required to complete all processes on each production line. The formula is expressed as: ; in, Indicates workpiece On the assembly line The Processing time on the machine. This represents the total number of processes, which is also the total number of machines on each production line. (2) The cumulative processing time of the workpiece in the first stage Second stage processing time Sort them separately; traverse each workpiece in order of its workpiece number. Choose to make the cumulative processing time of the first stage Minimum production line And a parallel machine that minimizes the processing time in the second stage. If it appears or In the same case, select the pipeline or parallel machine with the smaller number and generate a pipeline allocation scheme. Parallel machine allocation scheme ; (3) Based on the optimal pipeline selected in step (2) and optimal parallel machine Preliminary estimate of the total two-stage processing time for each workpiece. ,according to Sort all workpieces on the assembly line in ascending order, and generate the assembly line processing sequence using the shortest total processing time priority rule. If they are the same, sort them in ascending order by workpiece number.

[0030] The scheduling schemes and maximum completion times obtained from the heuristic greedy algorithm are shown in Table 2 below: Table 2 shows the solution results of the greedy algorithm.

[0031] As can be seen from the table, the heuristic greedy algorithm does solve the two-stage mixed job shop scheduling model. However, in the actual solution process, the greedy algorithm is prone to converge to a local optimum in the scheduling problem. In order to overcome this constraint and obtain a globally better scheduling scheme, this embodiment designs an improved genetic algorithm to solve the two-stage mixed job shop scheduling model, so as to obtain a better production scheduling scheme than the existing methods. Through the global search capability and iterative optimization mechanism of the improved genetic algorithm, it aims to significantly improve the operating efficiency of the production system and the comprehensive utilization rate of resources.

[0032] As a preferred embodiment, such as Figure 2 As shown, the improved genetic algorithm designed in this application is specifically as follows: Initialize the input information (i.e., input known, specific production data into the algorithm model), the input information including the workpiece set. Assembly line Assembly line process set Parallel machine set Processing time of each step on the assembly line Processing time of the workpiece on the parallel machine ; Construct a chromosome with a three-segment integer coding structure, and set the total length of each chromosome to the number of workpieces. Three times that of the previous three, with each encoded segment having a length of 1.5 times. The first segment of the code is a global sequential code, where the code position indicates the global order of workpiece processing in the assembly line, and the integer at each code position represents the workpiece number at the current position; the second segment of the code is a workpiece-assembly line allocation code, where the code position corresponds to the workpiece number, and the integer at each code position represents the assembly line number assigned to each workpiece; the third segment of the code is a workpiece-parallel machine allocation code, where the code position corresponds to the workpiece number, and the integer at each code position represents the parallel machine number assigned to the workpiece at the current position. Chromosome coding results as follows Figure 3 As shown, the global sequence of workpiece 1 can be read as 2. The first stage is assigned to pipeline 2, and the second stage is assigned to parallel machine 1.

[0033] The genetic algorithm parameters are determined, including population size, maximum number of iterations, crossover probability, and mutation probability. In this embodiment, the population size pop_size is set to 100, the maximum number of iterations n_gen is set to 50, and the crossover probability and mutation probability are 0.9 and 0.1, respectively.

[0034] Determine the termination condition and fitness function of the genetic algorithm; the termination condition is: the current iteration number is greater than the maximum iteration number; calculate the objective function value of each individual based on the scheduling result, and use it as the input of the fitness function. The lower the value of the objective function (i.e., the lower the maximum completion time), the better the scheduling scheme.

[0035] The initial population is divided into two categories, with each category containing half of the total population. The first category employs a random strategy: each segment of the three-segment integer encoding structure of each individual's chromosome is randomly generated. The first segment randomly generates the global order of workpiece processing in the first-stage assembly line. The second segment randomly assigns each workpiece to the assembly line according to the obtained global order. The third segment then randomly assigns each workpiece to a parallel machine. The second category employs the NEH+SPT strategy: first, according to the shortest time rule (SPT), each workpiece in the first-stage assembly line is assigned to the assembly line with the shortest total processing time, resulting in the second segment. In the second-stage parallel workshop, each workpiece is assigned to a parallel machine with the shortest processing time, resulting in the third segment. Finally, the total time for both stages is calculated, and the global order of workpiece processing in the first stage is generated using the NEH algorithm, resulting in the first segment.

[0036] In this embodiment, the NEH+SPT strategy adopted in the second type is specifically as follows: For each workpiece, calculate its total processing time on each production line. Select the production line that minimizes the cumulative processing time in the first stage for the workpiece. Parallel machines with the shortest processing time ; All workpieces are processed according to the total processing time in two stages. Sort in descending order, where... Get the initial list of workpieces to be inserted ; From the list of works to be inserted Remove the first two workpieces, generate all possible permutations, and evaluate the maximum completion time for each permutation. , retain The smaller sequence is used as the current processing sequence in the first stage. ; From the list of works to be inserted The next workpiece is taken out in sequence and inserted. From 0 to the sequence length Calculate the maximum completion time for each insertion scheme at all possible locations. Fix the workpiece in a position that minimizes Update the position ; Repeat the above steps until All workpieces are inserted into the sequence, forming the global order of workpiece processing in the first stage; at this point, the three-segment code of each individual chromosome is obtained; then, based on the scheduling results of the first stage, the completion time of the first stage for each workpiece is obtained. , press the workpiece By sorting the workpieces in ascending order and sequentially placing them onto the corresponding parallel machines for the second stage of processing, the processing scheduling scheme for the second stage can be obtained.

[0037] In this embodiment, different strategies are used to divide the initial population to improve solution efficiency while expanding the search range and increasing population diversity to avoid getting trapped in local convergence. The random strategy refers to the pipeline global order, pipeline allocation, and parallel machine allocation being completely randomly generated. While the quality of individuals in this type of population may not be high, it can cover different regions of the solution space, ensuring population diversity. The NEH+SPT heuristic strategy assigns higher priority to jobs with longer total processing times in both stages, iteratively adjusting the job order to find the processing sequence that minimizes pipeline bottlenecks and equipment waiting time. In the second stage, the shortest processing time rule is used, assigning each job to the parallel machine with the shortest processing time, generating a better initial solution than the random strategy. For subsequent population iterations, this allows the algorithm to explore the solution space more efficiently, improving convergence efficiency and solution quality.

[0038] Then, the individuals generated in each iteration are decoded to obtain a two-stage hybrid job shop scheduling scheme; production operations are carried out according to the obtained scheduling scheme, and the maximum completion time of the production task is calculated as the fitness of the current individual; In this embodiment, the decoding process specifically includes: First, read the chromosomes from left to right. The encoding of the bits is used to obtain the global order of workpiece processing in the first stage; based on the chromosome number... Ranked first The integer at each encoded position assigns a pipeline to each workpiece, and according to the workpiece sequence in the global processing order of the first stage, each workpiece is placed to the corresponding pipeline for processing, resulting in a production list for each pipeline. and the processing time matrix of each production line The production list This includes the workpiece processing sequence on each production line, and the processing time matrix. Including workpieces On the assembly line No. Processing time on the machine ; Initialize pipeline information and place the workpiece On the assembly line No. Start time, completion time on the machine, workpiece The initial values ​​for the completion time of the first stage are all set to zero; then according to and The completion time of each workpiece on each machine on the production line (i.e., the completion time of each process on the production line) is calculated recursively until the completion time of the first stage of each workpiece is obtained; it should be noted that the completion time in this application = start time + processing time. Read chromosome number again To the The bit encoding determines the workpiece-parallel machine allocation result, and the production list of each parallel machine is obtained. and the processing time matrix of each parallel machine The production list The processing time matrix includes the workpieces assigned to each parallel machine. This includes the processing time of each workpiece on the parallel machine; based on the completion time of each workpiece in the first stage, the actual processing order on each parallel machine is determined using a first-come, first-served rule (i.e., for multiple workpieces assigned to the same parallel machine, they are sorted in ascending order of their first-stage completion time to form the actual processing order on that parallel machine). Calculate the start time and finish time of each workpiece on the parallel machine, and recursively calculate the finish time of each parallel machine (i.e. the finish time of the last workpiece on the parallel machine) based on the actual processing order. Take the maximum value of the finish times of the parallel machines as the maximum finish time of the current production task.

[0039] The process involves determining whether the population iteration has reached the preset maximum number of iterations. If not, the optimal individual for the current iteration is saved based on fitness. Then, selection, crossover, and mutation operations are performed on the population. Specifically, after saving the optimal individual from the previous generation, individuals are selected from the subpopulation through a binary tournament to enter the mating pool and generate the next generation. Sequential crossover is used for the global sequential coding segment of the chromosome, while uniform crossover is used for the workpiece-pipeline allocation coding segment and the workpiece-parallel machine allocation coding segment. Random mutation is used for all three coding segments of the chromosome. After selection, crossover, and mutation operations, new individuals are obtained. The fitness of the new individuals is calculated and compared with the saved optimal individuals, and the optimal individuals are updated accordingly. If the maximum number of iterations is reached, the iteration stops, and the best individual that is finally saved is decoded to obtain the optimal scheduling scheme for the two-stage mixed operation workshop.

[0040] Based on the optimal production scheduling scheme, the workpiece-assembly line allocation result, workpiece-parallel machine allocation result, workpiece processing sequence and processing time plan are obtained, which guides the two-stage production resource scheduling of cabin parts and completes the production task.

[0041] This concludes the description of the method proposed in this application. Furthermore, this invention also discloses an electronic device comprising a memory and a processor, wherein: Memory is used to store computer programs that can run on a processor; The processor is configured to, while running the computer program, execute a two-stage mixed operation workshop scheduling method for cabin parts that takes into account cycle time differences, as described above.

[0042] Also disclosed is a computer-readable storage medium storing computer instructions for causing a processor to execute a two-stage mixed operation workshop scheduling method for cabin parts that takes into account cycle time differences, as described above.

[0043] The scheduling scheme and maximum completion time obtained by the method proposed in this invention are shown in Table 3 below: Table 3. Scheduling results of the method proposed in this invention

[0044] Table 3 clearly lists the resource allocation in the two stages, including the workpiece number to be processed, the corresponding processing sequence, and the completion time. The workpiece processing sequence is represented by a list of workpiece numbers, from left to right, indicating the processing order of the workpieces allocated to the corresponding resources. Based on the scheduling results above, the maximum completion time of the scheduling method proposed in this invention is 159 minutes, significantly better than the instance completion time obtained through existing heuristic greedy algorithms. This demonstrates that this embodiment effectively improves production efficiency through optimization using a genetic algorithm.

[0045] In summary, this invention proposes a two-stage hybrid operation workshop scheduling method for cabin parts that considers cycle time differences. Based on the production scenario, a two-stage hybrid operation workshop scheduling model is constructed with the objective function of minimizing the maximum completion time. A genetic algorithm is used to construct a three-segment code for efficient solution, effectively improving workshop production efficiency. This method fully considers the complex constraints of cycle time differences between processes and multi-line, multi-machine collaboration in actual production, achieving overall optimization of the two-stage scheduling of both assembly lines and parallel machines. This ensures the feasibility and balance of the scheduling scheme under multi-variety, small-batch production modes. The results of the embodiments show that the proposed method outperforms the traditional greedy strategy in terms of scheduling efficiency and resource utilization, significantly shortening the production cycle. It provides a theoretical basis and feasible technical support for handling multi-variety, small-batch production tasks in complex manufacturing fields such as aerospace, and has good engineering application prospects and promotional value.

[0046] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0047] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A two-stage hybrid job-shop scheduling method for cabin parts considering beat difference, characterized in that, Specifically comprising the following steps: Constructing a two-stage mixed job shop for cabin part production, collecting workpiece information and resource information of cabin part production; the two-stage mixed job shop is a mixed production system composed of a first-stage flow shop and a second-stage parallel shop; According to the collected workpiece information and resource information and production characteristics of the two-stage mixed job shop, a two-stage mixed job shop scheduling model is established, and a target function and production constraint conditions of the two-stage mixed job shop scheduling model are determined; An improved genetic algorithm is designed to solve the two-stage mixed job shop scheduling model and obtain an optimal production scheduling scheme; According to the optimal production scheduling scheme, workpiece-flow line allocation results, workpiece-parallel machine allocation results, workpiece processing sequence and processing time plan are obtained to guide two-stage production resource scheduling of cabin parts and complete the production task.

2. The method for two-stage hybrid job shop scheduling of cabin parts considering tempo difference according to claim 1, characterized in that, The workpiece information includes part production quantity and part processing procedure; the resource information includes the number of flow lines in the first stage, the number of machines on the flow lines, the processing time of each machine on the flow lines, the number of parallel machines in the second stage and the processing time of workpieces on each parallel machine.

3. The method for two-stage hybrid job shop scheduling of cabin parts considering tempo difference according to claim 1, characterized in that, The production characteristics of the two-stage mixed job shop specifically include: The first stage is a replacement flow shop, including a plurality of flow lines, each of which has a plurality of machines, representing a plurality of processes required for processing a workpiece a plurality of processes, the workpiece processing sequence of each machine on the same flow line being the same, and each workpiece being able to select only one flow line for processing and being unable to change the flow line each time. The second stage is a discrete parallel workshop, including The workpiece is selected flexibly to process the parallel machine. In production, each workpiece is required to pass through the two-stage job shop in sequence to complete the production task.

4. The method for two-stage hybrid job shop scheduling of cabin parts considering beat difference according to claim 3, characterized in that, The target function is to minimize the maximum completion time of the production task, and the maximum completion time of the production task is the maximum value of the completion times of the parallel machines in the second stage.

5. The method for two-stage hybrid job shop scheduling of cabin parts considering beat difference according to claim 3, characterized in that, The production constraint conditions include workpiece-flow line allocation uniqueness constraint, flow line procedure sequence constraint, two-stage connection constraint and parallel machine non-overlapping processing constraint; The workpiece-flow line allocation uniqueness constraint specifically is that in the first stage, each workpiece can be allocated to only one position of one flow line, and each position of each flow line can be occupied by only one workpiece; in the global sequence of first-stage workpiece processing, the workpiece at each position is allocated to only one flow line; The specific constraints on the production line process sequence are as follows: in the production line Above, position The completion time of each workpiece on each machine shall not be less than the processing time of that machine; position The workpiece in the machine The completion time is no earlier than when it is installed on the machine. Completion time, location of the workpiece In the machine The completion time of the workpiece is no earlier than that of the workpiece preceding it in the machine. Completion time; the location The first step in the global sequence of workpiece machining in the first stage One location; The two-stage connection constraint specifically is that in the second stage, each workpiece is allocated to only one parallel machine; the start time of the second stage is not earlier than the completion time of the first stage; The parallel machine non-overlapping processing constraint specifically is that for two workpieces allocated to the same parallel machine, the completion time of the workpiece processed first in the second stage is not later than the start time of the other workpiece in the second stage.

6. The method for two-stage hybrid job shop scheduling of cabin parts considering beat difference according to claim 1, characterized in that, The improved genetic algorithm for solving the two-stage mixed job shop scheduling model to obtain the optimal production scheduling scheme specifically includes: initializing input information, the input information including a set of workpieces , a set of flow lines , a set of flow line processes , a set of parallel machines , processing times of the workpieces at the processes of the flow lines , processing times of the workpieces at the parallel machines ; The total length of each chromosome is set as three times of the number of workpieces , and the length of each segment is , wherein the first segment is global sequence encoding, the encoding position represents the global sequence of the workpieces in the flow shop, and the integer on each encoding position represents the workpiece number at the current position; the second segment is workpiece-flow line assignment encoding, the encoding position corresponds to the workpiece number, and the integer on each encoding position represents the flow line number assigned to each workpiece; and the third segment is workpiece-parallel machine assignment encoding, the encoding position corresponds to the workpiece number, and the integer on each encoding position represents the parallel machine number assigned to the workpiece at the current position. Determining genetic algorithm parameters, including population size, maximum iteration number, crossover probability and mutation probability; Determining the termination condition and fitness function of the genetic algorithm; the termination condition is that the current iteration number is greater than the maximum iteration number; the target function value of each individual is calculated according to the scheduling result of each individual as the input of the fitness function; Divide the initial population into two categories, and the number of individuals in each category accounts for one half of the total number of individuals in the population; The first type adopts a random strategy: each segment in the three-segment integer coding structure of each individual chromosome in the first type is randomly generated, the first segment encodes the global sequence of the first stage flow shop job processing, the second segment encodes the random allocation of each job to the flow line according to the obtained global sequence, and the third segment encodes the random allocation of each job to the parallel machine; The second type adopts an NEH+SPT strategy: first, according to the shortest time rule SPT, allocate each job to the flow line with the shortest total processing time of each process in the first stage flow shop to obtain the second segment code; in the second stage parallel workshop, allocate each job to the parallel machine with the shortest processing time to obtain the third segment code; finally, calculate the total time of the two stages, generate the global sequence of the first stage job processing through the NEH algorithm, and obtain the first segment code; Decode the individual generated in each iteration to obtain a two-stage mixed job shop scheduling scheme; perform production work according to the obtained scheduling scheme, and calculate the maximum completion time of the production task as the fitness of the current individual; Determine whether the population iteration reaches the preset maximum iteration number, if not, save the optimal individual of the current iteration according to the fitness, then perform selection, crossover and mutation operations on the population to obtain a new individual; calculate the fitness of the new individual, compare it with the saved optimal individual, and update the optimal individual; If the maximum iteration number is reached, stop iteration, and decode the final saved optimal individual to obtain the optimal scheduling scheme of the two-stage mixed job shop.

7. The method for two-stage hybrid job shop scheduling of cabin parts considering beat difference according to claim 6, characterized in that, The decoding of the individual generated in each iteration to obtain a two-stage mixed job shop scheduling scheme specifically includes: First, read the chromosomes from left to right. The encoding of the bits is used to obtain the global order of workpiece processing in the first stage; based on the chromosome number... Ranked first The integer at each encoded position assigns a pipeline to each workpiece, and according to the workpiece sequence in the global processing order of the first stage, each workpiece is placed to the corresponding pipeline for processing, resulting in a production list for each pipeline. and the processing time matrix of each production line The production list This includes the workpiece processing sequence on each production line, and the processing time matrix. Including workpieces On the assembly line No. Processing time on the machine ; initializes the pipeline information, and sets the start time and the finish time of each job on each machine In the pipeline The first The start time and the finish time of each job on each machine are set to zero; then the finish time of each job on each machine is calculated recursively until the finish time of each job in the first stage is obtained The initial value of the finish time of each job in the first stage is set to zero; then the finish time of each job on each machine is calculated recursively until the finish time of each job in the first stage is obtained And The finish time of each job on each machine is calculated recursively until the finish time of each job in the first stage is obtained Read the chromosome again To the The bit encoding determines the workpiece-parallel machine allocation result, and the production list of each parallel machine is obtained. Processing time matrix of each parallel machine The production list The processing time matrix includes the workpieces assigned to each parallel machine. This includes the processing time of each workpiece on the parallel machine; According to the first stage completion time of each workpiece, the actual processing sequence on each parallel machine is determined by using the first-come-first-served rule, and the actual processing sequence is determined according to the first stage completion time of each workpiece and the first-come-first-served rule. The start time and completion time of each workpiece on the parallel machine are calculated, and the completion time of each parallel machine is recursively calculated based on the actual processing sequence, and the maximum value in the completion time of the parallel machine is taken as the maximum completion time of the current production task.

8. The method for two-stage hybrid job shop scheduling of cabin parts considering beat difference according to claim 6, characterized in that, The selection, crossover and mutation operations on the population to obtain a new individual specifically include: After saving the optimal individual of the previous generation, select individuals in the sub-population by binary tournament to enter the mating pool and generate the next generation; The crossover operation and the mutation operation are performed on the chromosome of the next generation individual, wherein: The crossover operation is: the global sequence coding segment of the chromosome adopts sequential crossover, and the job-flow line allocation coding segment and the job-parallel machine allocation coding segment adopt uniform crossover; The mutation operation is: random mutation is adopted for the three-segment coding of the chromosome; The new individual is obtained through the selection, crossover and mutation operations.

9. An electronic device, comprising: The electronic device includes a memory and a processor, wherein: The memory is used to store a computer program capable of running on the processor; The processor is used to execute the computer program when running, and performs a two-stage mixed job shop scheduling method for cabin parts considering beat difference according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, which are used to make the processor execute the two-stage mixed job shop scheduling method for cabin parts considering beat difference according to any one of claims 1-8 when executed.

Images