Processing and assembling combined production workshop scheduling optimization method considering dynamic disturbance

Abstract

The invention provides a processing and assembling combined production workshop scheduling optimization method considering dynamic disturbance, and the method comprises the steps: building a combined scheduling model: dividing a product production process into a part processing process, an assembling process and an assembling subsequent process, defining the correlation constraint of processing and assembling, and carrying out the processing and assembling operation; comprising part machining sequence consistency, part machining completion requirements before assembly, buffer area capacity limitation and a dynamic disturbance event response mechanism; establishing a multi-objective optimization function: taking minimization of maximum completion time, equipment idle time, processing energy consumption cost and tool switching times as optimization objectives, and generating a scheduling model in combination with process time dependence, machine selection uniqueness and tool switching time constraint; dynamic disturbance processing: when an equipment fault occurs or a new order is inserted, updating an available machine state or combining tasks to be processed, and triggering rescheduling to generate a new scheme; an improved MOEA / D algorithm is used for solving; and outputting and applying a scheduling scheme: selecting a final scheme for production guidance according to the non-dominated solution set.

Description

Technical Field

[0001] The present invention belongs to the technical fields of intelligent manufacturing, production scheduling optimization, etc., and particularly relates to a production workshop scheduling optimization method for joint processing and assembly considering dynamic disturbances. Background Art

[0002] Traditional production workshop scheduling research mainly focuses on the optimization of processing sequences and machine allocation for single processes. For example, in discrete manufacturing scenarios, heuristic rules (such as first-come, first-served, shortest processing time first) or static scheduling algorithms (such as genetic algorithms, tabu search) are used to determine the processing sequences and machine selections of each workpiece on independent processes. Such methods usually assume that there is no strong coupling relationship between processes, and the processing and assembly stages are treated separately - that is, all part processing tasks are completed first, and then subsequent operations are carried out based on a fixed assembly sequence. However, with the complexity of product structures (such as the multi-component assembly requirements of automobiles, aerospace vehicles, etc.), the coordination between part processing and assembly processes has become a key challenge:

[0003] Insufficient process relevance: Traditional scheduling models regard processing and assembly as independent stages and do not consider the impact of part processing sequences on assembly resource allocation. For example, if the processing of a certain part is delayed, it may lead to idle assembly processes or forced adjustment of assembly machine selections, thereby triggering a chain of inefficiencies in the overall scheduling plan;

[0004] Limitations in coping with dynamic disturbances: Existing research is mostly designed for static environments and lacks a real-time response mechanism for dynamic disturbances such as equipment failures and emergency order insertions. When disturbances occur, local rescheduling strategies (such as machine replacement for affected processes) are often adopted, but the multi-process association constraints are not globally optimized, easily leading to objective conflicts (such as the contradiction between minimizing the completion time and reducing energy consumption);

[0005] Deficiencies in multi-objective coordination: Although some research has introduced multi-objective optimization (such as NSGA-II), its objective functions are often limited to a single dimension of time or cost (such as the maximum completion time, machine load balancing), and it is difficult to balance the coordinated optimization of indicators such as efficiency, energy consumption, and equipment switching costs in complex scheduling scenarios.

[0006] In addition, in joint processing-assembly scheduling, practical constraints such as buffer capacity limitations and tooling changeover times further increase the complexity of the scheduling model, and the existing methods still have significant deficiencies in balancing such constraints and multi-objective coordination. Summary of the Invention

[0007] The present invention proposes a dynamic scheduling optimization method for joint processing and assembly for a production workshop that simultaneously includes part processing and assembly processes, aiming to solve the problems of the separation between the processing and assembly stages and the insufficient ability to respond to dynamic disturbances in traditional scheduling. Specifically, this method is implemented through the following technical framework:

[0008] Joint scheduling modeling: It is required that the workpiece sequences on all part processing machines are strictly consistent, and this sequence is used as the priority basis for selecting machines in the assembly process and subsequent processes to ensure the coordination of processes in the processing and assembly stages;

[0009] Dynamic disturbance response mechanism: For disturbance events such as equipment failures and new order insertions, the available status of machines or the set of tasks to be processed is updated in real time, triggering global rescheduling to quickly generate a feasible solution;

[0010] Multi-objective optimization solution: Based on the improved MOEA / D algorithm, combined with a three-layer coding structure, a dynamic weight vector update strategy, and deviation degree threshold control, efficient solution under multiple constraints is achieved.

[0011] The implementation process of the present invention includes the following steps:

[0012] Step S1: Formally describe the dynamic job shop scheduling problem for combined processing and assembly, and clarify the process division, disturbance types, and scheduling objectives;

[0013] Step S2: Define the key parameters and symbols of the scheduling model;

[0014] Step S3: Construct constraint conditions such as process time dependence, uniqueness of machine selection, and buffer capacity limit, and establish a multi-objective optimization function;

[0015] Step S4: Solve the model using the improved MOEA / D algorithm, including initializing the population, dynamically updating the weight vector, and managing the external archive to balance the solution stability before and after disturbances;

[0016] Step S5: Output the optimal scheduling plan that meets the constraint conditions to support manual decision-making and production execution.

[0017] The technical solution specifically adopted by the present invention to solve its technical problems is:

[0018] A production workshop scheduling optimization method for combined processing and assembly considering dynamic disturbances, including the following steps:

[0019] Construct a joint scheduling model: Divide the product production process into part processing processes, assembly processes, and subsequent assembly processes, and define the associated constraints for processing and assembly, including the consistency of part processing sequences, the requirement for part processing completion before assembly, buffer capacity limit, and dynamic disturbance event response mechanism;

[0020] Establish a multi-objective optimization function: With the objectives of minimizing the makespan, equipment idle time, processing energy consumption cost, and tooling changeover times, a scheduling model is generated in combination with process time dependence, uniqueness of machine selection, and tooling changeover time constraints;

[0021] Dynamic perturbation handling: When a device fails or a new order is inserted, update the available machine status or merge the tasks to be processed, and trigger rescheduling to generate a new plan;

[0022] Solving with the improved MOEA / D algorithm:

[0023] Generate an initial population using a three-layer coding structure, where the three-layer coding includes a part processing sequence layer, an assembly machine selection layer, and a subsequent process machine selection layer;

[0024] Update the weight vector through virtual objective vector expansion and crowding distance screening to drive the algorithm to converge;

[0025] Dynamically update the external archive based on a preset deviation degree threshold to balance the stability and optimality of the scheduling plan before and after perturbation; Output and apply the scheduling plan: Select the final plan according to the non-dominated solution set for production guidance.

[0026] Furthermore, the dynamic perturbation event response mechanism includes:

[0027] When a device fails, record the failed machine number and repair time, and reschedule the remaining workpieces after updating the machine available status;

[0028] When a new order is inserted, merge the new order with the unprocessed workpieces and re-optimize to ensure the constraints of the overall objective function.

[0029] Furthermore, the three-layer coding structure satisfies:

[0030] The part processing sequence layer constrains the consistency of the workpiece order of all processing machines;

[0031] The selection of assembly and subsequent process machines takes the part order as the priority, and maps the process assignment logic through a Gantt chart.

[0032] Furthermore, the division and constraints of the processes in the construction of the joint scheduling model include:

[0033] The part processing process is executed by a single machine, and the assembly process and the subsequent assembly processes are executed by multiple parallel machines;

[0034] The workpiece order on all part processing machines needs to be consistent, and this order is used as the priority basis for the selection of assembly processes and subsequent process machines;

[0035] Before the assembly process starts, the corresponding processing processes of all parts need to be completed, and a buffer is set between the assembly process and the subsequent processes. The buffer capacity is constrained to be allowed to enter only when the subsequent assembly process machines are idle.

[0036] Furthermore, the process time dependence constraint is specifically:

[0037] The start time of the subsequent process of the same workpiece shall not be earlier than the end time of the previous process;

[0038] When the machine continuously processes different types of workpieces, the tooling change time needs to be increased. The change time is preset according to the machine type and included in the process processing time.

[0039] Furthermore, when generating the initial population with the three-layer coding structure of the improved MOEA / D algorithm:

[0040] The first-layer coding represents the unified workpiece sequence on all part processing machines;

[0041] The second-layer coding represents the machine selection for the assembly process. According to the workpiece sequence of the first layer as the priority, multiple parallel machines are allocated;

[0042] The third-layer coding represents the machine selection for the subsequent assembly process, and the allocation logic is the same as that of the second-layer coding;

[0043] The coding length of a single individual is the sum of the number of workpieces in the three-layer coding, and the machine allocation and processing time sequence of each process are mapped through a Gantt chart.

[0044] Furthermore, the weight vector update strategy in the improved MOEA / D algorithm includes:

[0045] Generate an extended weight vector set, and calculate the minimum vertical projection distance between each extended vector and the current target vector set;

[0046] Filter the virtual target vectors according to the projection distance, and retain the vectors that meet the preset distance threshold;

[0047] Merge the sets through crowding distance sorting, and update the weight vectors to match the direction of the filtered target vectors.

[0048] Furthermore, the calculation of the deviation degree threshold includes:

[0049] Target deviation rate: the sum of the normalized differences between the rescheduling plan and the initial plan in terms of the makespan, equipment idle time, energy consumption cost, and the number of tooling changes;

[0050] Number of workpieces replaced on the equipment: the number of workpieces replaced on the equipment after rescheduling;

[0051] The deviation degree is the weighted sum result of the above two indicators, and the weight coefficient is a preset value.

[0052] Furthermore, the dynamic rescheduling external archive update mechanism includes:

[0053] When the number of solutions in the external archive solution set exceeds the upper limit, the solution with the largest deviation degree is preferentially deleted;

[0054] The solution generated by rescheduling needs to satisfy that the deviation degree is less than the preset threshold, otherwise secondary optimization is triggered.

[0055] Further, when the Gantt chart maps the process allocation logic:

[0056] The workpiece order of the assembly machine and the subsequent process machines inherits the priority of the first-layer coding;

[0057] When continuously processing different types of workpieces on the same machine, the tooling change time interval needs to be marked in the Gantt chart.

[0058] In addition, an electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the method described above are implemented.

[0059] A non-transitory computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, the steps of the method described above are implemented.

[0060] Compared with the prior art, the present invention and its preferred solutions have the following beneficial effects:

[0061] Process collaborative scheduling optimization: For a production workshop that simultaneously includes processing and assembly processes, a joint scheduling strategy is proposed, requiring the workpiece order on all part processing machines to be strictly consistent, and using this order as the priority basis for the assembly process and subsequent process to select machines, breaking the limitation of the separation between the processing and assembly stages in traditional scheduling, and significantly improving the process collaborative efficiency;

[0062] Dynamic disturbance response ability: For dynamic disturbance events such as equipment failures and new order insertions, a real-time rescheduling mechanism is designed. By updating the machine available status or merging the set of tasks to be processed, rapid response and global optimization after the disturbance occur are achieved, avoiding suboptimal solutions caused by local adjustments;

[0063] Improvement of multi-objective solving performance: Based on the improved MOEA / D algorithm, through a three-layer coding structure, a dynamic weight vector update strategy, and deviation degree threshold control, while ensuring the diversity of the solution set, the algorithm convergence is accelerated, effectively balancing the multi-objective optimization requirements of the makespan, equipment idle rate, energy consumption cost, and tooling change times. Description of the Drawings

[0064] The following further details the present invention in conjunction with the drawings and specific embodiments:

[0065] Figure 1 It is the method flowchart of the embodiment of the present invention;

[0066] Figure 2 It is the individual Gantt chart of the embodiment of the present invention. Detailed implementation manners

[0067] To make the features and advantages of this patent more obvious and understandable, specific embodiments are given below for detailed description as follows:

[0068] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in this specification have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0069] It should be noted that the terms used herein are only for describing specific implementation manners and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0070] As Figure 1 shown, this embodiment provides an optimization method for the production workshop scheduling of combined machining and assembly considering dynamic disturbances, and its construction process includes the following steps:

[0071] Step S1: Describe the production workshop scheduling problem of combined machining and assembly considering dynamic disturbances;

[0072] Step S2: Define relevant parameter symbols;

[0073] Step S3: Mathematically express relevant constraints and scheduling objectives;

[0074] Step S4: Solve the problem model using an improved MOEA / D algorithm;

[0075] Step S5: Output the scheduling plan.

[0076] In this embodiment, the implementation process of step S1 is specifically as follows:

[0077] The product production process is divided into C1 part processing processes, C2 assembly processes, and C3 post-assembly processes. Different from traditional job shops, the workshop involved in the present invention is a workshop for joint scheduling of processing and assembly. Among them, there are multiple parallel machines available for workpieces in the assembly process and the post-assembly process, and there is one machine for the remaining processes. The scheduling process involves arranging the workpiece sequence in the part processing process and arranging the workpiece sequence and machine selection in the assembly process and the post-assembly process. The processing sequence of parts is correlated with the assembly sequence, that is, it is required that the workpiece sequences corresponding to the parts processed on all part processing machines are consistent, and it is also required that the assembly process and the post-assembly process use this sequence as the priority order for selecting machines. Before the assembly process, all part processing processes need to be completed before assembly can be carried out. A buffer is provided between the assembly process and the post-assembly process.

[0078] The scheduling problem involved in the present invention also considers two disturbance factors of equipment failure and new order insertion during the production process. First, the initial order is scheduled to generate an initial scheduling plan. When an equipment failure occurs, the failed equipment is recorded, and at the same time, the equipment repair time is estimated. The available time of the failed equipment is updated in the machine set, and then the remaining unprocessed workpieces are counted for re-scheduling to generate a new scheduling plan. Similarly, if a new order needs to be inserted into production during the production process, the new order and the remaining unprocessed workpieces need to be summarized together for re-scheduling. The scheduling objectives are the makespan, equipment idle time, processing energy consumption cost, and tooling changeover times.

[0079] In this embodiment, in step S2, the definition of relevant parameter symbols is specifically as follows:

[0080] n The number of workpieces

[0081] m The total number of machines

[0082] N_T The number of objective functions

[0083] j The operation code

[0084] J The total number of operations

[0085] i The code of the workpiece

[0086] M The machine set, M = {1, 2, …, m}

[0087] k The machine number

[0088] S The operation set, S = {1, 2, …, J}

[0089] N The workpiece set, N = {1, 2, …, n}

[0090] i' The immediate predecessor workpiece of workpiece i

[0091] Oij The j-th process of workpiece i

[0092] A ik The product type of workpiece i processed on machine k

[0093] The rated processing power of the k-th machine

[0094] S ij The start time of processing the j-th process of workpiece i

[0095] E ij The end time of processing the j-th process of workpiece i

[0096] The start time of processing the j-th process of workpiece i on machine k

[0097] ST ij The time when the j-th process of workpiece i starts assembly

[0098] L The number of rescheduling times

[0099] P_T l The l-th rescheduling time

[0100] n_t l The set of unprocessed workpieces at the l-th rescheduling

[0101] R_t The time when the equipment fails

[0102] R The number of the failed equipment

[0103] T_R The estimated repair time of the failed equipment

[0104] The buffer entry time

[0105] BN The buffer capacity

[0106] t k The tool change time required for continuously processing different types of products on machine k

[0107] y j1,j2 Whether the j1-th process is a part of the j2-th process, 1 if yes, 0 otherwise

[0108] Whether the j-th process of workpiece i is processed on machine k, 1 if yes, 0 otherwise. Whether the process j of workpiece i processed on machine k and its machine-preceding workpiece are of the same type, if yes

[0109] v ijk is 0, otherwise 1. The first workpiece on machine k is also 0

[0110] u ij Whether the j-th process of workpiece i has a preceding process, if yes, it is 1, otherwise it is 0

[0111] In this embodiment, step S3 includes the following steps:

[0112] Step S31: Construct constraint conditions;

[0113] Step S32: Construct a scheduling objective function;

[0114] Furthermore, in step S31, the constraint conditions include:

[0115] 1) O ij Only one machine can be selected for processing:

[0116] 2) O ij Start and completion time relationship:

[0117] 3) The subsequent process can only be processed after the preceding process is completed:

[0118] S ij ≥E i,j-1 , j ∈ {2,…, J}.

[0119] 4) The tooling needs to be switched when the machine continuously processes different types of workpieces:

[0120]

[0121] 5) Before the assembly operation, all parts need to be processed before assembly can be carried out:

[0122] ST ij ≥ max(E ij' ·y j',j ), j ≠ j

[0123] 6) When the equipment fails, for the workpiece arranged to be processed on this machine, its start processing time is after the estimated repair time:

[0124]

[0125] 7) Buffer constraint, there must be vacancies in the buffer between the assembly process and the subsequent assembly process before assembly can be carried out.

[0126]

[0127] Furthermore, in step S32, the objective function includes:

[0128] 1) The maximum completion time f1:

[0129] f1 = max(E ij ),

[0130] 2) Equipment idle time f2:

[0131]

[0132] 3) Processing energy consumption cost f3:

[0133]

[0134] 4) Number of fixture changeovers f4:

[0135]

[0136] In this embodiment, step S4 is specifically as follows:

[0137] Step S41: Initialize the population. A three-layer coding method is designed according to the characteristics of the combined processing and assembly scheduling. The first layer is the processing sequence of each part, and it is required that the sequence of the workpieces corresponding to the processing machines of each part is the same; the second layer is the selection of the assembly machine, and the workpiece sequence of the first layer is used as the priority for selecting the machine; the third layer is the selection of the subsequent assembly process machine, and the workpiece sequence of the first layer is also used as the priority for selecting the machine. The number of workpieces is n, so the length of each layer is n. For example, when the number of workpieces is 6, there are 2 assembly machines and 3 subsequent assembly machines, the population size is generated as 100, and the coding length of a single individual is 6 for each of the 3 layers; as Figure 2 The Gantt chart for an individual [452316; 112122; 311232] is shown. The workpieces processed on each part machine are all 4, 5, 2, 3, 1; the machines are selected in this order, and the workpiece sequence on assembly machine 1 is 4, 5, 3; the workpiece sequence on assembly machine 2 is 2, 1, 6; the subsequent assembly machine sequence is the same.

[0138] Step S42: Randomly generate 100 weight vectors, denote the weight set as λ = {λ1, λ2,..., λ 100}, and calculate the scheduling objective function values by matching with the initial population, denote the objective vector set as W = {W1, W2,..., W 100};

[0139] Step S43: Weight vector update strategy;

[0140] Step S44: Dynamic rescheduling external archive update mechanism;

[0141] Step S45: Determine whether the maximum evaluation times have been reached. If so, output the optimal solution set; otherwise, update the offspring and then jump to step S44;

[0142] In this embodiment, step S43 includes the following steps:

[0143] Step S431, generating a virtual target vector: randomly regenerating a set of extended weight vectors λ with the number of vectors being N p . Then, calculate the virtual target vector by using the extended weight vector, and the calculation formula is as follows: p .

[0144]

[0145] Where

[0146]

[0147] For each weight vector λ in the set of extended weight vectors λ p , find the target vector f in the current target vector set W with the minimum distance d2 from it e , and generate the corresponding virtual target vector f through the direction of the extended weight vector and the position of the target vector m . Where f is the target vector after normalization; d2 is the distance from the target vector f to the perpendicular projection point P of the extended weight vector λ v ; d1 is the distance from P to the origin. e

[0148] Step S432, selection of representative target vectors: set a distance threshold μ. When the distance d2 between the virtual target vector and its corresponding target vector f m is less than μ, in this embodiment, preferably μ = 0.05, add this virtual target vector to the merged set W of the target vector set W a , then calculate the crowding distance of each vector in W a , sort them from smallest to largest, and retain the first 100 vectors, denoted as W e = {f e1 , f e2 ,... f e100}. The crowding distance calculation formula is as follows:

[0149]

[0150] Where LN i represents the i-th nearest Euclidean distance of the weight vector, and N_T is the number of objective functions. In this embodiment, the number of objective functions is 4.

[0151] Step S433, update the weight vector through the following formula:

[0152]

[0153] In this embodiment, step S44: When there is no disturbance, the non-dominated solution set is normally updated to the external archive. At the end of each scheduling, it is necessary for a person to select a scheduling plan as the production guidance. When a disturbance occurs and rescheduling is performed, the deviation degree between the new scheduling plan and the previous production plan needs to be calculated when updating the external archive. When the external archive exceeds the upper limit, the solution with a large deviation degree is removed. The calculation formula for the deviation degree S_D is as follows:

[0154]

[0155] where C_D i represents the deviation rate of the four objective function values. M_D represents the number of workpieces replaced by the equipment after rescheduling. ε1 = 0.5, ε2 = 0.5.

[0156] Based on the same inventive concept, the present invention also provides a computer device, which includes: one or more processors, and a memory for storing one or more computer programs; the program includes program instructions, and the processor is used to execute the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is used to implement one or more instructions, specifically for loading and executing one or more instructions in the computer storage medium to implement the above method.

[0157] It should be further noted that, based on the same inventive concept, the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is run by a processor, it executes the above method. The storage medium may be any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electrical, magnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium may be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0158] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second", and similar terms used in the present invention do not denote any order, quantity, or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or items appearing before the term cover the elements or items listed after the term and their equivalents, without excluding other elements or items. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0159] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention in any other form. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

[0160] This patent is not limited to the above-mentioned best implementation mode. Anyone inspired by this patent can come up with various other forms of a production workshop scheduling optimization method for combined processing and assembly considering dynamic disturbances. All equivalent changes and modifications made within the scope of the patent application according to this invention shall fall within the scope covered by this patent.

Claims

1. A production workshop scheduling optimization method for combined processing and assembly considering dynamic disturbances, characterized in that: The following steps are involved: Construct a joint scheduling model: divide the product production process into parts processing process, assembly process and assembly post-process, and define the associated constraints between processing and assembly, including the consistency of parts processing sequence, the requirements for parts processing completion before assembly, the buffer capacity limit and the dynamic disturbance event response mechanism; Establish a multi-objective optimization function: minimize the maximum completion time, equipment idle time, processing energy cost and tooling switching times as the optimization goals, and generate a scheduling model by combining process time dependency, machine selection uniqueness and tooling switching time constraints; Dynamic disturbance processing: When equipment fails or new orders are inserted, the available machine status is updated or the tasks to be processed are merged, triggering rescheduling to generate new plans; Improved MOEA / D algorithm solution: An initial population is generated using a three-layer coding structure, wherein the three-layer coding includes a part processing sequence layer, an assembly machine selection layer, and a subsequent process machine selection layer; Update the weight vector through virtual target vector expansion and crowding distance screening to drive the algorithm convergence; Dynamically update external archives based on preset deviation thresholds to balance the stability and optimization of scheduling plans before and after disturbances; Output and apply the scheduling plan: Select the final plan for production guidance based on the non-dominated solution set.

2. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The dynamic disturbance event response mechanism includes: When equipment fails, record the faulty machine number and repair time, update the machine availability status and reschedule the remaining workpieces; When a new order is inserted, the new order is merged with the unprocessed workpiece and then re-optimized to ensure the overall objective function constraints.

3. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The three-layer coding structure satisfies: The part processing sequence layer constrains the workpiece sequence consistency of all processing machines; Machine selection for assembly and subsequent operations is prioritized based on part sequence, and the operation allocation logic is mapped via a Gantt chart.

4. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The division and constraints of the processes in constructing the joint scheduling model include: The parts processing process is performed by a single machine, and the assembly process and subsequent assembly processes are performed by multiple parallel machines; The order of workpieces on all parts processing machines must be consistent, and this order is used as the priority basis for machine selection for assembly and subsequent processes; Before the assembly process begins, the processing procedures of all corresponding parts must be completed, and a buffer zone must be set between the assembly process and subsequent processes. The buffer zone capacity is constrained so that the machine is only allowed to enter when the subsequent assembly process is idle.

5. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The process time dependency constraints are specifically: The start time of the subsequent process of the same workpiece shall not be earlier than the end time of the previous process; When the machine continuously processes different types of workpieces, it is necessary to increase the tooling switching time. The switching time is preset according to the machine type and is included in the process processing time.

6. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: When the three-layer coding structure of the improved MOEA / D algorithm generates the initial population: The first level of coding represents a uniform workpiece sequence on all part processing machines; The second level of coding represents the machine selection for the assembly process, which is assigned to multiple parallel machines based on the priority of the first level workpiece sequence; The third level of coding indicates the machine selection for the subsequent assembly process, and the allocation logic is consistent with the second level of coding; The coding length of a single individual is the sum of the number of workpieces in the three-level coding, and the machine allocation and processing timing of each process are mapped through a Gantt chart.

7. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The weight vector update strategy in the improved MOEA / D algorithm includes: Generate an extended weight vector set, and calculate the minimum vertical projection distance between each extended vector and the current target vector set; Filter virtual target vectors according to the projection distance and retain vectors that meet a preset distance threshold; The collection is merged by sorting by crowding distance and the weight vector is updated to match the direction of the filtered target vector.

8. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The calculation of the deviation degree threshold includes: Target deviation rate: the sum of the normalized differences between the rescheduling plan and the initial plan in terms of maximum completion time, equipment idle time, energy consumption cost, and number of tooling switching times; Number of workpieces replaced by equipment: the number of workpieces replaced on the equipment after rescheduling; The degree of deviation is the weighted sum of the above two indicators, and the weight coefficient is the preset value.

9. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: The dynamic rescheduling external file update mechanism includes: When the number of external archive solution sets exceeds the upper limit, the solutions with the largest deviation are deleted first; The solution generated by rescheduling must satisfy the deviation degree less than the preset threshold, otherwise secondary optimization will be triggered.

10. The method for optimizing production workshop scheduling for combined processing and assembly considering dynamic disturbances according to claim 1, characterized in that: When mapping the process allocation logic in the Gantt chart: The workpiece sequence of the assembly machine and the subsequent process machine inherits the priority of the first level coding; When different types of workpieces are processed continuously on the same machine, the time interval for tooling switching needs to be marked in the Gantt chart.

Images