Multi-stage collaborative scheduling method for ship block assembly workshop

Abstract

The invention relates to a multi-stage collaborative scheduling method for a ship block assembly workshop, and belongs to the technical field of assembly workshop scheduling. The ship block assembly job shop scheduling modeling is an assembly job shop collaborative scheduling problem model considering three stages of processing, transportation and assembly; and solving an assembly job shop collaborative scheduling problem model of three stages of processing, transportation and assembly based on an improved genetic algorithm. Taking minimization of completion time as an optimization target, and converting a part integrity constraint in a machining stage, a time difference constraint of a part in a transportation stage and a flowing water blockage constraint in an assembly stage into control variable constraints in a mathematical model; the strategy is improved on the basis of the genetic algorithm, and the solving capacity of the genetic algorithm for the ship block assembly job shop scheduling problem is improved; the method effectively solves the problem of ship block assembly workshop scheduling, and has solving precision and stability.

Description

Technical Field

[0001] The invention relates to a multi-stage collaborative scheduling method for an assembly workshop, and belongs to the technical field of scheduling of assembly workshops. Background Art

[0002] The assembly workshop scheduling problem is a classic scheduling problem, the core of which is that each part follows a specific processing route in the processing stage and is assembled into the final product in the assembly stage. This type of problem is solved by designing a scheduling scheme under the constraints of workshop resources to optimize one or more scheduling indicators. A high-quality scheduling scheme can not only improve the utilization rate of various resources, but also effectively shorten the assembly cycle. For example: the publication number is CN106651049A, and the invention is named a rescheduling method for automated container terminal loading and unloading equipment. Its technical scheme optimizes the scheduling scheme of each device in the three stages; then after a time period, the state monitoring controller measures the actual position and actual speed of each device in the three stages; updates the minimum completion time required for an upcoming task according to the state measurement values ​​of the actual position and actual speed; then, the state monitoring controller monitors the updated task time according to the minimum completion time required for an upcoming task; finally, the state monitoring controller determines a new scheduling scheme for each device in the three interacting stages; the present invention has the advantage of providing a fast and accurate rescheduling scheme when the loading and unloading equipment is disturbed. However, compared with general scheduling problems, the scheduling of assembly workshops is more complicated, and the scheduling optimization of ship block assembly workshops is even more complicated, requiring overall planning of workshop operations in the three stages of processing, transportation, and assembly. In the optimization process, it is necessary to consider the parts integrity constraints in the processing stage, the time difference constraints of parts in the transportation stage, and the blockage constraints in the assembly stage. In view of the above multi-stage constraints, it is of great significance to study the scheduling method suitable for ship block assembly workshops and formulate efficient scheduling plans to achieve cost reduction and efficiency improvement in ship block assembly workshops.

[0003] Prior art document 1 "An Improved Elephant Herding Optimization for Energy-Saving Assembly Job Shop Scheduling Problem with Transportation Times", Jiang et al., Axioms, 2022, 11(10): 23. This document aims at the assembly job shop scheduling problem and takes minimizing energy consumption as the goal. It proposes a meta-heuristic solution algorithm based on improved elephant herding. However, this document does not consider the form of multi-workstation flow layout in the assembly stage; Technical document 2 "Inventory and total completion time minimization for assembly job-shop scheduling considering material integrity and assembly sequential constraint", Cheng et al., Journal of Manufacturing Systems, 2022, 65: 660-672. This document aims at the assembly job shop scheduling problem and takes minimizing job completion time and inventory as the goal. It proposes a meta-heuristic solution algorithm based on improved simulated annealing. However, this document does not consider the time difference constraints of different parts in the transportation stage and the congestion constraints between workstations in the assembly stage.

[0004] Therefore, it is urgent to propose a multi-stage collaborative scheduling method for ship block assembly workshops to solve the above technical problems. Summary of the invention

[0005] In order to solve the above problems, a multi-stage collaborative scheduling method for a ship block assembly workshop is provided. A brief overview of the invention is given below to provide a basic understanding of certain aspects of the invention. It should be understood that this overview is not an exhaustive overview of the invention. It is not intended to identify the key or important parts of the invention, nor is it intended to limit the scope of the invention.

[0006] The technical solution of the present invention:

[0007] A multi-stage collaborative scheduling method for a ship block assembly workshop comprises the following steps:

[0008] Step 1: Model the scheduling of the ship block assembly workshop as a collaborative scheduling problem model of the assembly workshop taking into account the three stages of processing-transportation-assembly. Taking the minimization of completion time as the optimization goal, the parts integrity constraint in the processing stage, the time difference constraint of the parts in the transportation stage, and the blocked flow constraint in the assembly stage are transformed into control variable constraints in the mathematical model; specifically, the following steps are included:

[0009] Construct an assembly job shop scheduling optimization objective with the goal of minimizing completion time;

[0010] Transform the part integrity constraints in the processing stage into control variable constraints in the mathematical model;

[0011] Transform the transportation time difference constraints in the transportation stage into control variable constraints in the mathematical model;

[0012] Transform the blocked flow constraints in the assembly phase into control variable constraints in the mathematical model;

[0013] Step 2: Based on the characteristics of the ship block assembly workshop scheduling model, a variety of improvement strategies are designed on the basis of the genetic algorithm to enhance the genetic algorithm's ability to solve the ship block assembly workshop scheduling problem. The resulting scheduling scheme can effectively solve the ship block assembly workshop scheduling problem with high solution accuracy and stability; specifically, the following steps are included:

[0014] S1: population initialization;

[0015] S2: Chromosome crossover operation;

[0016] S3: Chromosome mutation operation;

[0017] S4: fitness value calculation;

[0018] S5: Termination judgment of algorithm optimization;

[0019] The present invention is a method applied to the workshop scheduling problem of ship block assembly operation.

[0020] Preferably, in step 1, the assembly workshop scheduling optimization objective of minimizing the completion time is constructed, specifically:

[0021]

[0022] Among them, f is the optimization target, P is the segment set, p is the segment index value, B p represents the completion time of the segment corresponding to p.

[0023] Preferably, in step 1, the part integrity constraint in the processing stage is converted into a control variable constraint in a mathematical model, which is expressed as any segmented part. Only when all processes of the part are completed can the transportation stage be carried out, specifically:

[0024]

[0025] Among them, J p represents the set of processed parts used to assemble the segment corresponding to p, j is the index value of the processed part, O jp For processing parts collection J p Where j is the process set of the part corresponding to it, o is the process index value, | jp | is the process set O jp The last process index value in the assembly phase, W is the workstation set in the assembly phase, w is the workstation index value, and b jpw Represents the state variable. When the processing part set J p When the part corresponding to j is assembled on the workstation corresponding to w, then b jpw =1, otherwise b jpw =0, is the process set O jp The process completion time corresponding to o in the figure is: for b jpw = 1 when j corresponds to the start of part transportation, Indicates any.

[0026] Preferably: in step 1, the transportation time difference constraint of the transportation stage is converted into a mathematical model control variable constraint, which is expressed as follows: the transportation time of each part depends on the type of parts in the segment, and the premise for the workstation to carry out assembly work is that the required parts have been transported to the designated workstation, specifically:

[0027]

[0028] Where T is the assembly segment sequence set, t is the assembly segment sequence set index value, τ jp Represents the processing parts set J p The parts transportation time corresponding to j is, It indicates the time when the segment corresponding to t starts to be assembled on the workstation corresponding to w.

[0029] Preferably, in step 1, the blocking flow constraint in the assembly stage is converted into a mathematical model control variable constraint, which is expressed as:

[0030] The precondition for the start of the assembly operation of a segment at a workstation is that the segment has completed the assembly operation of the previous workstation and there is no segment staying at the workstation. Specifically:

[0031]

[0032] The precondition for the segment transfer event between workstations is that the segment has completed the assembly operation at the current workstation and there is no segment stop at the next workstation. Specifically:

[0033]

[0034] There is no blocking between workstations in the first assembly segment, and there is no blocking in the last assembly workstation; specifically:

[0035]

[0036] Among them, N is the maximum index value of the workstation, I is the maximum index value of the assembly segment sequence set T, represents the end time of the assembly segment corresponding to t at the workstation corresponding to w, represents the end time of the first assembly segment on the workstation corresponding to w, It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to w-1 (the previous workstation of the workstation corresponding to w). It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to N (the last workstation); represents the departure time of the assembly segment corresponding to t at the workstation corresponding to w, represents the departure time of the first assembly segment at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 (the preceding segment of the assembly segment corresponding to t) at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 at the workstation corresponding to w+1, represents the departure time of the assembly segment corresponding to t at the workstation corresponding to N.

[0037] Preferably, the S1: population initialization specifically includes the following steps:

[0038] S11, chromosome coding design:

[0039] A double-layer coded chromosome structure is designed to clearly represent the complex relationship between processing and assembly. The first layer of coding is aimed at the segmented assembly operation, called assembly coding, which adopts a non-repeating integer coding form, where each integer represents a segment, and the assembly workstations perform assembly operations in sequence according to the integer coding order in the chromosome. The second layer of coding is aimed at the segmented processing operation, called processing coding, which adopts the coding form of a two-dimensional vector (p, j), where p is the segment index value and j is the processing part index value. The same two-dimensional vector in the chromosome represents the same processing part in the same segment, and the number of occurrences represents the operation order of the processing part. Based on the above coding ideas, a double-layer chromosome coding structure suitable for the workshop scheduling problem of ship segmented assembly operations is constructed.

[0040] S12, decision parameter setting:

[0041] Considering the large difference in the solution space scale between assembly coding and processing coding, a number of decision parameters are designed to improve the algorithm's solving ability; the specific parameters include: the maximum number of algorithm iterations, population size, assembly crossover probability, assembly mutation probability, processing crossover probability, processing mutation probability, neighborhood search distance, and the number of experimental repetitions for each group;

[0042] S13, initial population construction:

[0043] Due to the strong coupling between the assembly and processing stages, it is usually difficult to generate a better scheduling solution by simply using random initialization; therefore, a hybrid initial population generation method combining assembly-driven initialization strategy and random initialization is adopted; the steps of the assembly-driven initialization strategy are as follows: first, the assembly code of the chromosome is randomly initialized to determine the assembly order of the segments; then, based on the generated assembly code, the processing code of the chromosome is driven to be generated according to the assembly order, so as to implement the strategy of processing the segmented parts that are assembled first first; the steps of the random initialization strategy are to randomly initialize the assembly code and processing code of the chromosome respectively.

[0044] Preferably: the S2: chromosome crossover operation specifically comprises the following steps:

[0045] S21, assembly coding cross operation:

[0046] The two parent chromosomes used for the crossover operation are represented as P1 and P2, and the two daughter chromosomes generated after the crossover are represented as O1 and O2. Each chromosome contains two parts: assembly code and processing code. First, two position index values ​​are randomly selected within the length range of the assembly code of P1 or P2. In order to ensure that the generated offspring is a feasible solution, the gene values ​​within the index value range in the P1 assembly code are first directly copied to the corresponding positions of the O1 assembly code. Secondly, each gene value in the P2 assembly code is retrieved from front to back. When the O1 assembly code does not contain the gene value, the gene value is inherited to the O1 assembly code, otherwise the gene value is skipped and the next gene value is retrieved. Similarly, the offspring O2 assembly code is generated in the same way.

[0047] Due to the strong coupling between the assembly and processing stages, a processing code exchange operation based on assembly drive is designed in the process of inheriting the processing codes of P1 and P2 to O1 and O2. The specific steps are as follows: First, the gene values ​​of the same gene positions of the P1 assembly code and the O1 assembly code are compared to form a gene mapping rule; second, based on the rule, the P1 processing code is exchanged. For example, if the mapping rule of a certain gene position is 3→4, all the processing parts corresponding to segments 3 and 4 in the P1 processing code are exchanged, and the exchanged results are inherited to the processing code of O1. Similarly, the processing code of the offspring O2 is generated by this method, and finally a complete offspring chromosome is formed;

[0048] S22, processing code cross operation:

[0049] Under different assembly codes, the same processing code exhibits different fitness values. Therefore, two chromosomes with the same assembly code are selected as parents P1 and P2, and a crossover operation is performed on their processing codes. The main operation flow is similar to the crossover operation of the assembly code in S21, but there are the following differences: in S21, the crossed assembly code is a non-repeating integer sequence, while in the processing code, there are repeated gene values ​​to represent the operation order of the processing operation. Therefore, in the process of retrieving the complementary gene value, if the maximum operation order of the processing part has been reached in the processing code of O1, the complementary operation of the gene value is abandoned and the next gene value is retrieved, and finally the crossover operation of the processing code is realized and the daughter chromosomes O1 and O2 are produced.

[0050] Preferably: the S3: chromosome mutation operation specifically includes:

[0051] S31, assembly encoding mutation operation:

[0052] Randomly select an assembly node from the set of job nodes on the critical path, randomly select another index node within the neighborhood of the assembly code, exchange the positions of the two nodes, and perform the assembly-driven processing code exchange operation in S21 on the processing code of the same chromosome;

[0053] S32, processing code variation operation:

[0054] An index node is randomly selected within the processing coding neighborhood, and another index node is randomly selected within the processing coding neighborhood, and the positions of the two nodes are exchanged, thereby realizing the chromosome mutation operation.

[0055] Preferably, the step S4: calculating the fitness value specifically includes:

[0056] A semi-active decoding strategy is used to calculate the fitness value of each chromosome, and the process is divided into two stages. The first stage is to decode the processing operation sequence formed by the processing code. In this process, based on the processing machines assigned to each process, when the processing machines and the process sequence meet the processing requirements, processing begins, otherwise, continue to wait, and calculate the completion time of all parts in sequence. The second stage is to decode the assembly operation sequence formed by the assembly code. In this stage, based on the assembly sequence of each segment, when the parts supply, assembly workstation and assembly sequence meet the assembly requirements, assembly begins, otherwise, continue to wait, and calculate the assembly completion time of all segments in sequence. Finally, the maximum value of the assembly completion time of all segments is the fitness value of the chromosome.

[0057] Preferably, the termination judgment of the algorithm optimization in S5 specifically includes:

[0058] Determine whether the number of iterations has been reached, if so, output the optimal solution set, if not, jump to step S2;

[0059] The algorithm uses two stopping strategies to work together. As long as one of the stopping strategies is met, the algorithm stops calculating.

[0060] Stop strategy 1: The maximum number of algorithm iterations is 200;

[0061] Stopping strategy 2: If no new optimal solution appears after more than 100 consecutive iterations, the algorithm converges.

[0062] The present invention has the following beneficial effects:

[0063] The present invention models the scheduling of ship block assembly workshops as a collaborative scheduling problem model of assembly workshops that takes into account the three stages of processing, transportation and assembly. The optimization goal is to minimize the completion time. The parts integrity constraints in the processing stage, the time difference constraints of the parts in the transportation stage, and the blocked flow constraints in the assembly stage are converted into control variable constraints in the mathematical model, and the time difference constraints of different parts in the transportation stage and the blocking constraints between workstations in the assembly stage are fully considered.

[0064] The invention is based on the characteristics of the ship block assembly workshop scheduling model, designs multiple improvement strategies based on the genetic algorithm, and enhances the genetic algorithm's ability to solve the ship block assembly workshop scheduling problem; the invention effectively solves the ship block assembly workshop scheduling problem and has high solution accuracy and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 The present invention is a flow chart of a multi-stage collaborative scheduling method for a ship block assembly workshop.

[0066] Figure 2 It is a Gantt chart for the workshop scheduling of ship block assembly operations;

[0067] Among them, the A-axis represents the assembly time in minutes, and the B-axis represents the processing machines (M1-M10) and assembly workstations (W1-W5);

[0068] Figure 3 It is a box diagram of the scheduling results of a ship block assembly workshop formed by the multi-stage collaborative scheduling method of the ship block assembly workshop and other methods;

[0069] The B-axis represents the assembly time in min. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is described below by the specific embodiments shown in the accompanying drawings. However, it should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0071] Specific implementation method 1: Combination Figure 1-3 This embodiment is described. A multi-stage collaborative scheduling method for a ship block assembly workshop in this embodiment includes the following steps:

[0072] Step 1: Model the scheduling of the ship block assembly workshop as a collaborative scheduling problem model of the assembly workshop taking into account the three stages of processing-transportation-assembly; specifically, the following steps are included:

[0073] Construct an assembly workshop scheduling optimization goal with the goal of minimizing completion time; complete all tasks in the shortest possible time, thereby improving production efficiency and reducing delivery delays;

[0074] The part integrity constraints in the processing stage are converted into control variable constraints in the mathematical model; each part meets the requirements during the processing;

[0075] Transform the transportation time difference constraints in the transportation stage into control variable constraints in the mathematical model; effectively avoid the time difference from affecting the overall progress, better optimize the transportation process, and facilitate management;

[0076] Transform the blocked flow constraints in the assembly phase into control variable constraints in the mathematical model; represent and manage them through control variables in the mathematical model to ensure the smooth progress of the assembly process;

[0077] Step 2: Solve the collaborative scheduling problem model of the assembly workshop in the three stages of processing, transportation and assembly based on the improved genetic algorithm; specifically, the following steps are included:

[0078] S1: population initialization;

[0079] S2: Chromosome crossover operation;

[0080] S3: Chromosome mutation operation;

[0081] S4: fitness value calculation;

[0082] S5: Termination judgment of algorithm optimization.

[0083] Specific implementation method 2: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is used. In step 1, the assembly workshop scheduling optimization objective of minimizing the completion time is constructed, specifically:

[0084]

[0085] Among them, f is the optimization target, P is the segment set, p is the segment index value, B p represents the completion time of the segment corresponding to p.

[0086] Specific implementation method three: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. In step 1, the part integrity constraint in the processing stage is converted into a control variable constraint in a mathematical model, which is expressed as any segmented part. Only when all processes of the part are completed can the transportation operation be carried out. Specifically,

[0087]

[0088] Among them, J p represents the set of processed parts used to assemble the segment corresponding to p, j is the index value of the processed part, O jp For processing parts collection J p Where j is the process set of the part corresponding to it, o is the process index value, | jp | is the process set O jp The last process index value in the assembly phase, W is the workstation set in the assembly phase, w is the workstation index value, and b jpw Represents the state variable. When the processing part set J p When the part corresponding to j is assembled on the workstation corresponding to w, then b jpw =1, otherwise b jpw =0, is the process set O jp The process completion time corresponding to o in the figure is: for b jpw = 1 when j corresponds to the start of part transportation, Indicates any.

[0089] Specific implementation method four: Combination Figure 1-3 This embodiment is described. This embodiment is a multi-stage collaborative scheduling method for a ship block assembly workshop. In step 1, the transportation time difference constraint of the transportation stage is converted into a mathematical model control variable constraint, which is expressed as the transportation time of each part depends on the type of parts in the section. The premise for the workstation to carry out assembly work is that the required parts have been transported to the designated workstation, specifically:

[0090]

[0091] Where T is the assembly segment sequence set, t is the assembly segment sequence set index value, τ jp Represents the processing parts set J p The parts transportation time corresponding to j is, It indicates the time when the segment corresponding to t starts to be assembled on the workstation corresponding to w.

[0092] Specific implementation method five: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. In step 1, the blocked flow constraint of the assembly stage is converted into a mathematical model control variable constraint, which is expressed as:

[0093] The precondition for the start of the assembly operation of a segment at a workstation is that the segment has completed the assembly operation of the previous workstation and there is no segment staying at the workstation. Specifically:

[0094]

[0095] The precondition for the segment transfer event between workstations is that the segment has completed the assembly operation at the current workstation and there is no segment stop at the next workstation. Specifically:

[0096]

[0097] There is no blocking between workstations in the first assembly segment, and there is no blocking in the last assembly workstation; specifically:

[0098]

[0099] Among them, N is the maximum index value of the workstation, I is the maximum index value of the assembly segment sequence set T, represents the end time of the assembly segment corresponding to t at the workstation corresponding to w, represents the end time of the first assembly segment on the workstation corresponding to w, It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to w-1 (the previous workstation of the workstation corresponding to w). It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to N (the last workstation); represents the departure time of the assembly segment corresponding to t at the workstation corresponding to w, represents the departure time of the first assembly segment at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 (the preceding segment of the assembly segment corresponding to t) at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 at the workstation corresponding to w+1, represents the departure time of the assembly segment corresponding to t at the workstation corresponding to N.

[0100] Specific implementation method six: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. S1: population initialization specifically includes the following steps:

[0101] S11, chromosome coding design:

[0102] A double-layer coded chromosome structure is designed to clearly represent the complex relationship between processing and assembly. The first layer of coding is aimed at the segmented assembly operation, called assembly coding, which adopts a non-repeating integer coding form, where each integer represents a segment, and the assembly workstations perform assembly operations in sequence according to the integer coding order in the chromosome. The second layer of coding is aimed at the segmented processing operation, called processing coding, which adopts the coding form of a two-dimensional vector (p, j), where p is the segment index value and j is the processing part index value. The same two-dimensional vector in the chromosome represents the same processing part in the same segment, and the number of occurrences represents the operation order of the processing part. Based on the above coding ideas, a double-layer chromosome coding structure suitable for the workshop scheduling problem of ship segmented assembly operations is constructed.

[0103] S12, decision parameter setting:

[0104] Considering the large difference in the solution space scale between assembly coding and processing coding, a number of decision parameters are designed to improve the algorithm's solving ability; the specific parameters include: the maximum number of algorithm iterations, population size, assembly crossover probability, assembly mutation probability, processing crossover probability, processing mutation probability, neighborhood search distance, and the number of experimental repetitions for each group;

[0105] S13, initial population construction:

[0106] Due to the strong coupling between the assembly and processing stages, it is usually difficult to generate a better scheduling solution by simply using random initialization; therefore, a hybrid initial population generation method combining assembly-driven initialization strategy and random initialization is adopted; the steps of the assembly-driven initialization strategy are as follows: first, the assembly code of the chromosome is randomly initialized to determine the assembly order of the segments; then, based on the generated assembly code, the processing code of the chromosome is driven to be generated according to the assembly order, so as to implement the strategy of processing the segmented parts that are assembled first first; the steps of the random initialization strategy are to randomly initialize the assembly code and processing code of the chromosome respectively.

[0107] Specific implementation method seven: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. The S2: chromosome crossover operation specifically includes the following steps:

[0108] S21, assembly coding cross operation:

[0109] The two parent chromosomes used for the crossover operation are represented as P1 and P2, and the two daughter chromosomes generated after the crossover are represented as O1 and O2. Each chromosome contains two parts: assembly code and processing code. First, two position index values ​​are randomly selected within the length range of the assembly code of P1 or P2. In order to ensure that the generated offspring is a feasible solution, the gene values ​​within the index value range in the P1 assembly code are first directly copied to the corresponding positions of the O1 assembly code. Secondly, each gene value in the P2 assembly code is retrieved from front to back. When the O1 assembly code does not contain the gene value, the gene value is inherited to the O1 assembly code, otherwise the gene value is skipped and the next gene value is retrieved. Similarly, the offspring O2 assembly code is generated in the same way.

[0110] Due to the strong coupling between the assembly and processing stages, a processing code exchange operation based on assembly drive is designed in the process of inheriting the processing codes of P1 and P2 to O1 and O2. The specific steps are as follows: First, the gene values ​​of the same gene positions of the P1 assembly code and the O1 assembly code are compared to form a gene mapping rule; second, based on the rule, the P1 processing code is exchanged. For example, if the mapping rule of a certain gene position is 3→4, all the processing parts corresponding to segments 3 and 4 in the P1 processing code are exchanged, and the exchanged results are inherited to the processing code of O1. Similarly, the processing code of the offspring O2 is generated by this method, and finally a complete offspring chromosome is formed;

[0111] S22, processing code cross operation:

[0112] Under different assembly codes, the same processing code exhibits different fitness values. Therefore, two chromosomes with the same assembly code are selected as parents P1 and P2, and a crossover operation is performed on their processing codes. The main operation flow is similar to the crossover operation of the assembly code in S21, but there are the following differences: in S21, the crossed assembly code is a non-repeating integer sequence, while in the processing code, there are repeated gene values ​​to represent the operation order of the processing operation. Therefore, in the process of retrieving the complementary gene value, if the maximum operation order of the processing part has been reached in the processing code of O1, the complementary operation of the gene value is abandoned and the next gene value is retrieved, and finally the crossover operation of the processing code is realized and the daughter chromosomes O1 and O2 are produced.

[0113] Specific implementation method eight: Combination Figure 1-3 This embodiment is described. This embodiment is a multi-stage collaborative scheduling method for a ship block assembly workshop. The S3: chromosome mutation operation specifically includes:

[0114] S31, assembly encoding mutation operation:

[0115] Randomly select an assembly node from the set of job nodes on the critical path, randomly select another index node within the neighborhood of the assembly code, exchange the positions of the two nodes, and perform the assembly-driven processing code exchange operation in S21 on the processing code of the same chromosome;

[0116] S32, processing code variation operation:

[0117] An index node is randomly selected within the processing coding neighborhood, and another index node is randomly selected within the processing coding neighborhood, and the positions of the two nodes are exchanged, thereby realizing the chromosome mutation operation.

[0118] Specific implementation method nine: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. The S4: fitness value calculation specifically includes:

[0119] A semi-active decoding strategy is used to calculate the fitness value of each chromosome, and the process is divided into two stages. The first stage is to decode the processing operation sequence formed by the processing code. In this process, based on the processing machines assigned to each process, when the processing machines and the process sequence meet the processing requirements, processing begins, otherwise, continue to wait, and calculate the completion time of all parts in sequence. The second stage is to decode the assembly operation sequence formed by the assembly code. In this stage, based on the assembly sequence of each segment, when the parts supply, assembly workstation and assembly sequence meet the assembly requirements, assembly begins, otherwise, continue to wait, and calculate the assembly completion time of all segments in sequence. Finally, the maximum value of the assembly completion time of all segments is the fitness value of the chromosome.

[0120] Specific implementation method ten: Combination Figure 1-3 This embodiment is described. In this embodiment, a multi-stage collaborative scheduling method for a ship block assembly workshop is described. The termination judgment of the algorithm optimization in S5 specifically includes:

[0121] Determine whether the number of iterations has been reached, if so, output the optimal solution set, if not, jump to step S2;

[0122] The algorithm uses two stopping strategies to work together. As long as one of the stopping strategies is met, the algorithm stops calculating.

[0123] Stop strategy 1: The maximum number of algorithm iterations is 200;

[0124] Stopping strategy 2: If no new optimal solution appears after more than 100 consecutive iterations, the algorithm converges.

[0125] Embodiment 1:

[0126] Firstly, based on the business characteristics of the ship block assembly workshop, the present invention models the scheduling of the ship block assembly workshop as a collaborative scheduling problem model of the assembly workshop taking into account the three stages of processing-transportation-assembly, and realizes a unified mathematical model description of the parts integrity constraint in the processing stage, the transportation time difference constraint in the transportation stage, and the blocked flow constraint in the assembly stage;

[0127] Step 1: Constructing the collaborative scheduling model of the assembly workshop in the three stages of processing, transportation and assembly

[0128] The ship section is assembled from multiple parts, and the processing of each part includes multiple processes. In the part processing stage, each part completes each process on a designated processing machine in a specific processing sequence. After the processing is completed, the part is transferred to a buffer zone, and when the assembly conditions of the part are met, it is transported to a designated assembly workstation. The section passes through each workstation on the assembly line in a predetermined sequence to perform specific assembly operations. The present invention aims to minimize the completion time of the ship section assembly workshop by optimizing the scheduling scheme, so as to improve the overall manufacturing efficiency.

[0129] The ship block assembly workshop scheduling problem model consists of two parts: objective function and constraints:

[0130] Taking minimizing the completion time as the optimization goal, the objective function is:

[0131]

[0132] Among them, f is the optimization target, P is the segment set, p is the segment index value, B p represents the completion time of the segment corresponding to p;

[0133] Constraints: Convert the part integrity constraints in the processing stage into control variable constraints of the mathematical model:

[0134]

[0135] Among them, J p represents the set of processed parts used to assemble the segment corresponding to p, j is the index value of the processed part, O jp For processing parts collection J p Where j is the process set of the part corresponding to it, o is the process index value, | jp | is the process set O jp The last process index value in the assembly process; W is the workstation set in the assembly stage, and w is the workstation index value; b jpw Represents the state variable. When the processing part set J p When the part corresponding to j is assembled on the workstation corresponding to w, then b jpw =1, otherwise b jpw=0; is the process set O jp The process completion time corresponding to o in the figure is: for b jpw = 1, the corresponding part transportation time of j; Formula (2) indicates that for any segmented part, transportation can be carried out only when all processes of the part are completed;

[0136] The transportation time difference constraints in the transportation stage are transformed into control variable constraints of the mathematical model:

[0137]

[0138] Where T is the assembly segment sequence set, and t is the assembly segment sequence set index value. jp Represents the processing parts set J p The parts transportation time corresponding to j is, represents the time when the segment corresponding to t starts assembly at the workstation corresponding to w; Formula (3) indicates that the transportation time of each part depends on the part type of the segment, and the premise for the workstation to carry out assembly work is that the required parts have been transported to the designated workstation;

[0139] The blocking flow constraints in the assembly stage are converted into control variable constraints of the mathematical model:

[0140]

[0141] Among them, N is the maximum index value of the workstation, I is the maximum index value of the assembly segment sequence set T, represents the end time of the assembly segment corresponding to t at the workstation corresponding to w, represents the end time of the first assembly segment on the workstation corresponding to w, It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to w-1 (the previous workstation of the workstation corresponding to w). It indicates the end time of the assembly segment corresponding to t at the workstation corresponding to N (the last workstation); represents the departure time of the assembly segment corresponding to t at the workstation corresponding to w, represents the departure time of the first assembly segment at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 (the preceding segment of the assembly segment corresponding to t) at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t-1 at the workstation corresponding to w+1, represents the departure time of the assembly segment corresponding to t at the workstation corresponding to N. Formula (4) indicates that the premise for the segment to start assembling at a workstation is that the segment has completed the assembly at the previous workstation and no segment has stayed at the workstation; Formula (5) indicates that the premise for the segment to transfer between workstations is that the segment has completed the assembly at the current workstation and no segment has stayed at the next workstation; Formula (6) indicates that there is no blocking between workstations for the first assembly segment, and Formula (7) indicates that there is no blocking in the last assembly workstation;

[0142] At this point, a unified mathematical model description has been developed for the parts integrity constraints in the processing stage, the transportation time difference constraints in the transportation stage, and the blocked flow constraints in the assembly stage, realizing the construction of a collaborative scheduling problem model for assembly workshops taking into account the three stages of processing-transportation-assembly.

[0143] Secondly, the present invention solves the collaborative scheduling problem model of the assembly workshop with three stages of processing, transportation and assembly based on an improved genetic algorithm; the algorithm adopts a double-layer coding structure to clearly express the complex relationship between processing and assembly, and enhances the algorithm's solving ability by setting multiple decision variables; in order to improve the quality and diversity of the initial population, the algorithm adopts a mixed population initialization strategy; in addition, through the assembly-driven strategy, the coupling relationship between assembly and processing is optimized, thereby minimizing the part delay phenomenon, and the critical path-driven domain search strategy is adopted to further enhance the algorithm's ability to jump out of the local optimal solution, thereby more effectively exploring potential scheduling configuration solutions;

[0144] Step 2: Solve the collaborative scheduling problem model of the assembly workshop in the three stages of processing-transportation-assembly based on the improved genetic algorithm

[0145] The solution to the ship block assembly workshop scheduling problem based on the improved genetic algorithm includes the following steps:

[0146] S1: Population initialization. In the population initialization stage, there are three operations: chromosome coding design, decision parameter setting, and initial population construction;

[0147] S11: Chromosome coding design. The present invention designs a double-layer coded chromosome structure to clearly represent the complex relationship between processing and assembly. The first layer of coding is for segmented assembly operations, called assembly coding, and adopts a non-repeating integer coding form, where each integer represents a segment, and the assembly workstations perform assembly operations in sequence according to the integer coding order in the chromosome. The second layer of coding is for segmented processing operations, called processing coding, and adopts a two-dimensional vector (p, j) coding form, where p is the segment index value and j is the processing part index value. The same two-dimensional vector in the chromosome represents the same processing part of the same product, and the number of occurrences represents the operation order of the processing part. Based on the above coding idea, the present invention constructs a double-layer chromosome coding structure suitable for the workshop scheduling problem of ship segmented assembly operations. For example, the present invention implements chromosome coding construction based on the double-layer coding structure design rules, such as assembly coding AS = {3, 2, 1, 4, ...}, processing coding PS = {(3, 1), (2, 1), (3, 1) ...}.

[0148] S12: Decision parameter setting. Considering the large difference in solution space scale between assembly coding and processing coding, the present invention designs a number of decision parameters to improve the algorithm's solution capability. The specific parameters include: the maximum number of algorithm iterations, population size, assembly crossover probability, assembly mutation probability, processing crossover probability, processing mutation probability, neighborhood search distance, and the number of experimental repetitions per group. For example, the maximum number of algorithm iterations is 200, the population size is 200, the assembly crossover probability is 0.5, the assembly mutation probability is 0.1, the processing crossover probability is 0.5, the processing mutation probability is 0.3, the neighborhood search distance is 2, and the number of experimental repetitions per group is 10.

[0149] S13: Initial population construction. Due to the strong coupling between the assembly and processing stages, it is usually difficult to generate a better scheduling solution by simply using random initialization. Therefore, the present invention adopts a hybrid initial population generation method that combines an assembly-driven initialization strategy with random initialization. The steps of the assembly-driven initialization strategy are as follows: First, the assembly code of the chromosome is randomly initialized to determine the assembly order of the segments. Then, based on the generated assembly code, the processing code of the chromosome is driven to be generated according to the assembly order, so as to implement the strategy of processing the segmented parts that are assembled first first. The steps of the random initialization strategy are to randomly initialize the assembly code and the processing code of the chromosome respectively. For example, the present invention constructs the initial population according to the hybrid initial population generation method, and the ratio of the number of chromosomes generated by the assembly-driven initialization strategy and the random initialization strategy is 0.2:0.8.

[0150] S2: Chromosome crossover operation, in the chromosome crossover stage, including assembly coding crossover and processing coding crossover;

[0151] S21: Assembly code crossover operation. When the assembly code satisfies the crossover probability, the assembly code crossover operation is performed, and the processing codes belonging to the same chromosome are exchanged;

[0152] The two parent chromosomes used for the crossover operation of the present invention are represented as P1 and P2, and the two daughter chromosomes generated after the crossover are represented as O1 and O2, and each chromosome includes two parts: an assembly code and a processing code; firstly, two position index values ​​are randomly selected within the length range of the assembly code of P1 or P2, and in order to ensure that the generated offspring is a feasible solution, the gene values ​​within the index value range in the P1 assembly code are first directly copied to the corresponding positions of the O1 assembly code, and then each gene value in the P2 assembly code is retrieved from front to back, and when the O1 assembly code does not contain the gene value, the gene value is inherited to the O1 assembly code, otherwise the gene value is skipped and the next gene value is retrieved; similarly, the offspring O2 assembly code is generated according to the same method;

[0153] Due to the strong coupling between the assembly and processing stages, the present invention designs a processing code exchange operation based on assembly drive in the process of inheriting the processing codes of P1 and P2 to O1 and O2; the specific steps are as follows: first, the gene values ​​of the same gene positions of the P1 assembly code and the O1 assembly code are compared to form a gene mapping rule; second, based on the rule, the P1 processing code is exchanged. For example, if the mapping rule of a certain gene position is 3→4, all the processing parts corresponding to segments 3 and 4 in the P1 processing code are exchanged, and the exchanged result is inherited to the processing code of O1. Similarly, the processing code of the offspring O2 is generated by this method, and finally a complete offspring chromosome is formed;

[0154] S22: Processing code crossover operation, when the processing code satisfies the crossover probability, two chromosomes with the same assembly code are selected for crossover operation;

[0155] Under different assembly codes, the same processing code exhibits different fitness values. Therefore, the present invention selects two chromosomes with the same assembly code as the parent generations P1 and P2, and performs a crossover operation on their processing codes. The main operation flow is similar to the crossover operation of the assembly code in S21, but there are the following differences: in S21, the crossed assembly code is a non-repeating integer sequence, while in the processing code, there are repeated gene values ​​to represent the operation order of the processing operation. Therefore, in the process of retrieving the complementary gene value, if the maximum operation order of the processing part has been reached in the processing code of O1, the complementary operation of the gene value is abandoned and the next gene value is retrieved, and finally the crossover operation of the processing code is realized and the daughter chromosomes O1 and O2 are produced.

[0156] S3: Chromosome mutation operation. Considering that the scheduling problem of the ship block assembly operation workshop has multimodal characteristics and the meta-heuristic algorithm is prone to fall into the local optimal solution, the present invention designs a chromosome mutation operation based on the critical path neighborhood search; in the chromosome mutation stage, it includes two operations: assembly code mutation and processing code mutation;

[0157] S31: assembly code mutation operation, randomly select an assembly node from the set of job nodes on the critical path, randomly select another index node within the assembly code neighborhood, swap the positions of the two nodes, and perform the assembly-driven processing code swap operation in S21 on the processing code of the same chromosome;

[0158] S32: Processing code mutation operation. When the chromosome meets the processing mutation probability, randomly select an index node within the processing code neighborhood, and randomly select another index node within the processing code neighborhood, and exchange the positions of the two nodes, thereby realizing the chromosome mutation operation.

[0159] S4: Fitness value calculation, based on the decoding strategy, calculate the fitness value of each chromosome in the population;

[0160] The present invention adopts a semi-active decoding strategy to calculate the fitness value of each chromosome, and the process is divided into two stages. In the first stage, the processing operation sequence formed by the processing code is decoded. In this process, based on the processing machine assigned to each process, when the processing machine and the process sequence meet the processing requirements, the processing starts, otherwise, it continues to wait and calculates the completion time of all parts in sequence. In the second stage, the assembly operation sequence formed by the assembly code is decoded. In this stage, based on the assembly sequence of each segment, when the parts supply, assembly workstation and assembly sequence meet the assembly requirements, the assembly starts, otherwise, it continues to wait and calculates the assembly completion time of all segments in sequence. Finally, the maximum value of the assembly completion time of all segments is the fitness value of the chromosome.

[0161] S5: termination judgment of the algorithm optimization, judging whether the number of iterations has been reached, if so, outputting the optimal solution set, if not, jumping to the step S2;

[0162] The algorithm uses two stopping strategies to work together. As long as one of the stopping strategies is met, the algorithm stops calculating.

[0163] Stop strategy 1: The maximum number of algorithm iterations is 200;

[0164] Stopping strategy 2: If no new optimal solution appears after more than 100 consecutive iterations, the algorithm converges;

[0165] The example of the present invention is derived from the abstraction of actual production cases. This example includes 8 segmented manufacturing operations, each segment contains a maximum of 10 parts, each part contains a maximum of 10 processes, involving 10 processing machines (M1-M10) and 5 assembly workstations (W1-W5). Based on the optimal solution obtained, a scheduling plan for the ship segment assembly operation workshop is formed, such as Figure 2 As shown, Figure 2 A Gantt chart of ship block assembly workshop scheduling drawn after solving the method of the present invention;

[0166] At this point, the entire process of the multi-stage collaborative scheduling method for a ship block assembly workshop of the present invention is completed;

[0167] The present invention models the scheduling of the ship block assembly workshop as a collaborative scheduling problem model of the assembly workshop taking into account the three stages of processing-transportation-assembly, takes the minimization of completion time as the optimization goal, and transforms the part integrity constraint in the processing stage, the transportation time difference constraint in the transportation stage, and the blocked flow constraint in the assembly stage into the control variable constraint in the mathematical model; secondly, the present invention solves the scheduling problem of the ship block assembly workshop based on the improved genetic algorithm. The algorithm uses a double-layer coding structure to represent the complex relationship between processing and assembly, sets multiple decision variables to enhance the algorithm's solving ability, adopts a mixed population initialization strategy to improve the quality and diversity of the initial population, and optimizes the coupling relationship between assembly and processing through an assembly-driven strategy, thereby minimizing the part delay phenomenon, and adopts a critical path-driven domain search strategy to further enhance the algorithm's ability to jump out of the local optimal solution, thereby more effectively exploring potential scheduling configuration schemes; the ship block assembly operation scheduling scheme obtained by the present invention can effectively shorten the assembly operation period.

[0168] In order to verify the beneficial effects of the method of the present invention, the method of the present invention is compared with other methods for verification; wherein IEHO is the solution algorithm of the technical document 1 partially cited in the technical field of the present invention, and ISA is the solution algorithm of the technical document 2 partially cited in the technical field of the present invention; the GA-TS solution algorithm is derived from the technical document "An effective hybrid genetic algorithm and tabu search for flexible job shop scheduling problem", Li et al., International Journal of Production Economics, 2016, 174: 93-110, which has strong superiority in solving the workshop job scheduling problem; the comparison of the box plots of the method of the present invention and the above-mentioned method shows that the method of the present invention has better solution accuracy and stability;

[0169] In summary, the multi-stage collaborative scheduling method of a ship block assembly workshop in the present invention is more in line with the actual operation background of the ship block assembly workshop, and the proposed improved genetic algorithm has strong applicability. It can be seen that the method proposed in the present invention is of great significance for guiding the scheduling of ship block assembly workshops.

[0170] It should be noted that in the above embodiments, as long as the technical solutions are not contradictory, they can be arranged and combined, and those skilled in the art can exhaust all possibilities based on the mathematical knowledge of arrangement and combination. Therefore, the present invention will no longer describe the technical solutions after arrangement and combination one by one, but it should be understood that the technical solutions after arrangement and combination have been disclosed by the present invention.

[0171] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-stage collaborative scheduling method for a ship block assembly workshop, characterized in that: The following steps are involved: The scheduling model of the ship block assembly workshop is a collaborative scheduling problem model of the assembly workshop considering the three stages of processing-transportation-assembly; it specifically includes the following steps: Construct an assembly job shop scheduling optimization objective with the goal of minimizing completion time; Transform the part integrity constraints in the processing stage into control variable constraints in the mathematical model; Transform the transportation time difference constraints in the transportation stage into control variable constraints in the mathematical model; Transform the blocked flow constraints in the assembly phase into control variable constraints in the mathematical model; The collaborative scheduling problem model of the assembly workshop in the three stages of processing, transportation and assembly is solved based on the improved genetic algorithm; the specific steps include: S1: population initialization; S2: chromosome crossover operation; S3: Chromosome mutation operation; S4: fitness value calculation; S5: Termination judgment of algorithm optimization.

2. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 1, characterized in that: The assembly workshop scheduling optimization objective of minimizing the completion time is constructed as follows: Among them, f is the optimization target, P is the segment set, p is the segment index value, B p represents the completion time of the segment corresponding to p.

3. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 2, characterized in that: The part integrity constraints in the processing stage are converted into control variable constraints in the mathematical model, specifically: Among them, J p represents the set of processed parts used to assemble the segment corresponding to p, j is the index value of the processed part, O jp For processing parts collection J p Where j is the process set of the part corresponding to it, o is the process index value, | jp | is the process set O jp The last process index value in the assembly phase, W is the workstation set in the assembly phase, w is the workstation index value, and b jpw Represents the state variable. When the processing part set J p When the part corresponding to j is assembled on the workstation corresponding to w, then b jpw =1, otherwise b jpw =0, is the process set O jp The process completion time corresponding to o in the figure is: for b jpw =1, the time when j corresponding to the start of part transportation.

4. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 3, characterized in that: The transportation time difference constraint of the transportation stage is converted into the control variable constraint of the mathematical model, specifically: Where T is the assembly segment sequence set, t is the assembly segment sequence set index value, τ jp Represents the processing parts set J p The parts transportation time corresponding to j is, It indicates the time when the segment corresponding to t starts to be assembled on the workstation corresponding to w.

5. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 4, characterized in that: The blocking flow constraints in the assembly stage are converted into control variable constraints of the mathematical model, which can be expressed as: The precondition for the start of the assembly operation of a segment at a workstation is that the segment has completed the assembly operation of the previous workstation and there is no segment staying at the workstation. Specifically: The precondition for the segment transfer event between workstations is that the segment has completed the assembly operation at the current workstation and there is no segment stop at the next workstation. Specifically: There is no blocking between workstations in the first assembly segment, and there is no blocking in the last assembly workstation; specifically: Among them, N is the maximum index value of the workstation, I is the maximum index value of the assembly segment sequence set T, represents the end time of the assembly segment corresponding to t at the workstation corresponding to w, represents the departure time of the assembly segment corresponding to t at the workstation corresponding to w.

6. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 5, characterized in that: The S1: population initialization specifically includes the following steps: S11, chromosome coding design: A double-layer coded chromosome structure is designed to represent the complex relationship between processing and assembly. The first layer of coding is for segmented assembly operations, called assembly coding, which uses non-repeating integer coding. Each integer represents a segment, and the assembly workstations perform assembly operations in sequence according to the integer coding order in the chromosome. The second level of coding is for segmented processing operations, called processing coding, which uses a two-dimensional vector (p, j) encoding format, where p is the segment index value and j is the processing part index value; the same two-dimensional vector in the chromosome represents the same processing part in the same segment, and the number of occurrences indicates the operation order of the processing part; S12, decision parameter setting: Specific parameters include: maximum number of algorithm iterations, population size, assembly crossover probability, assembly mutation probability, processing crossover probability, processing mutation probability, neighborhood search distance, and number of experimental repetitions for each group; S13, initial population construction: A hybrid initial population generation method combining assembly-driven initialization strategy and random initialization is adopted; the steps of the assembly-driven initialization strategy are as follows: randomly initialize the assembly code of the chromosome and determine the assembly order of the segments; based on the generated assembly code, drive the generation of the chromosome's processing code according to the assembly order to implement the strategy of processing the segmented parts that are assembled first first; the steps of the random initialization strategy are to randomly initialize the assembly code and processing code of the chromosome respectively.

7. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 6, characterized in that: The S2: chromosome crossover operation specifically includes the following steps: S21, assembly coding cross operation: The two parent chromosomes used for the crossover operation are represented as P1 and P2, and the two daughter chromosomes generated after the crossover are represented as O1 and O2. Each chromosome contains two parts: assembly code and processing code. Two position index values ​​are randomly selected within the length range of the assembly code of P1 or P2. In order to ensure that the generated offspring is a feasible solution, the gene value within the index value range in the P1 assembly code is directly copied to the corresponding position of the O1 assembly code, and each gene value in the P2 assembly code is retrieved from front to back. When the O1 assembly code does not contain the gene value, the gene value is inherited to the O1 assembly code, otherwise the gene value is skipped and the next gene value is retrieved. Similarly, the offspring O2 assembly code is generated according to the same method. Due to the strong coupling between the assembly and processing stages, a processing code exchange operation based on assembly drive is designed in the process of inheriting the processing codes of P1 and P2 to O1 and O2; the specific steps are as follows: compare the gene values ​​of the same gene positions of the P1 assembly code and the O1 assembly code to form a gene mapping rule; based on the rule, the P1 processing code is exchanged. For example, if the mapping rule of a certain gene position is 3→4, all the processing parts corresponding to segments 3 and 4 in the P1 processing code are exchanged, and the exchanged results are inherited to the processing code of O1. Similarly, the processing code of the offspring O2 is generated through this method, and finally a complete offspring chromosome is formed; S22, processing code cross operation: Under different assembly codes, the same processing code exhibits different fitness values, so two chromosomes with the same assembly code are selected as parents P1 and P2, and a crossover operation is performed on their processing codes.

8. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 7, characterized in that: The S3: chromosome mutation operation specifically includes: S31, assembly encoding mutation operation: Randomly select an assembly node from the set of job nodes on the critical path, randomly select another index node within the neighborhood of the assembly code, exchange the positions of the two nodes, and perform the assembly-driven processing code exchange operation in S21 on the processing code of the same chromosome; S32, processing code variation operation: An index node is randomly selected within the processing coding neighborhood, and another index node is randomly selected within the processing coding neighborhood, and the positions of the two nodes are exchanged, thereby realizing the chromosome mutation operation.

9. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 8, characterized in that: The S4: fitness value calculation specifically includes: A semi-active decoding strategy is used to calculate the fitness value of each chromosome, and the process is divided into two stages. The first stage is to decode the processing operation sequence formed by the processing code. In this process, based on the processing machines assigned to each process, when the processing machines and the process sequence meet the processing requirements, processing begins, otherwise, continue to wait, and calculate the completion time of all parts in sequence. The second stage is to decode the assembly operation sequence formed by the assembly code. In this stage, based on the assembly sequence of each segment, when the parts supply, assembly workstation and assembly sequence meet the assembly requirements, assembly begins, otherwise, continue to wait, and calculate the assembly completion time of all segments in sequence. The maximum value of the assembly completion time of all segments is the fitness value of the chromosome.

10. A multi-stage collaborative scheduling method for ship block assembly workshops according to claim 9, characterized in that: The termination judgment of S5: algorithm optimization specifically includes: Determine whether the number of iterations has been reached, if so, output the optimal solution set, if not, jump to step S2; The algorithm uses two stopping strategies to work together. If one of the stopping strategies is met, the algorithm stops calculating. Stop strategy 1: The maximum number of algorithm iterations is 200; Stopping strategy 2: If no new optimal solution appears after more than 100 consecutive iterations, the algorithm converges.

Images