A method and system for planning a parallel disassembly sequence of multiple waste oil-immersed transformers considering batch scheduling

Abstract

This invention provides a method and system for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling, belonging to the field of green dismantling and recycling of decommissioned power equipment. Addressing the problems of workstation load imbalance, bufferless congestion flow interference, batch flow mutual influence, and optimization difficulties under high-dimensional constraints in the batch dismantling of multiple scrap oil-immersed transformers, this invention, while adhering to multi-dimensional composite constraints and time domain boundary conditions, establishes a temporal and spatial scheduling matrix of batch and workstation two-dimensional indices for the arrival, operation, and departure times of each transformer at each workstation. Based on the constraints of batch occupancy before and after at the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, a bufferless congestion flow recursive completion time model for multiple transformers across multiple workstations is constructed. A multi-objective optimization solution is then performed based on an improved genetic algorithm, outputting a batch dismantling scheduling scheme containing information on the arrival, operation, and departure times of each transformer at each workstation, and generating a visualized Gantt chart.

Description

Technical Field

[0001] This invention relates to the field of green dismantling and recycling technology for decommissioned power equipment, and more particularly to a method and system for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling. Background Technology

[0002] Oil-immersed power transformers, as key equipment in power systems, have complex internal structures, containing high-value materials such as copper windings, iron cores, and silicon steel sheets. With the upgrading of power equipment, a large number of old transformers are entering their end-of-life stage. Their dismantling and recycling process involves multiple steps, such as draining oil, removing the outer casing, and disassembling the windings. Due to the large size and weight of transformers, the dismantling process relies on lifting equipment, specialized tools, and multi-station collaborative operations. Traditional dismantling methods often rely on manual experience and lack systematic process planning, resulting in low dismantling efficiency, especially when processing multiple transformers in batches, where uneven task allocation between workstations and low resource utilization are particularly prominent. Furthermore, transformer dismantling is subject to strict process constraints, such as certain components requiring dismantling at specific workstations, and the sequential dependencies between some processes, further increasing the complexity of batch dismantling scheduling.

[0003] Existing research on batch dismantling mainly focuses on small electronic products or automotive parts, where the dismantling process is relatively simple and typically assumes an infinite buffer between workstations, allowing semi-finished products to accumulate freely. However, this assumption is not applicable to transformer dismantling. Due to the large size of transformers, insufficient buffer space between workstations prevents timely transfer of products from upstream to downstream workstations, causing congestion and severely impacting dismantling efficiency. Furthermore, existing dismantling planning methods often employ a serial dismantling model, completing dismantling one unit at a time before moving on to the next, failing to fully utilize the advantages of multi-workstation parallel operation, resulting in excessively long overall completion times. Even with a parallel dismantling model, the lack of a scientific batch scheduling strategy leads to workstation load imbalances, with some workstations idle while others are overloaded.

[0004] At the algorithmic level, traditional genetic algorithms perform poorly when dealing with high-dimensional complex constraint problems, such as topology priority constraints, workstation compatibility constraints, and batch scheduling constraints. Their standard crossover and mutation operations cannot guarantee the generation of feasible solutions, requiring additional computational resources for repair, and they are prone to getting trapped in local optima. In addition, existing methods lack sufficient support for visualization of the decomposition progress, making it difficult to intuitively display information such as workstation occupancy and task progress, which is not conducive to real-time adjustment and optimization of on-site operations.

[0005] Therefore, there is an urgent need for a dismantling sequence planning method that can comprehensively consider batch scheduling, workstation load balancing and congestion flow mechanism to improve the efficiency and feasibility of batch dismantling multiple scrap oil-immersed transformers. Summary of the Invention

[0006] Based on the aforementioned technical problems, existing technologies fail to incorporate the batch dismantling of multiple machines into a unified collaborative planning paradigm, lack global planning leading to workstation load imbalance and low resource utilization, traditional methods ignore the physical space constraints of no intermediate buffer area causing congestion flow interference, and the solution space of the optimization model for the batch dismantling sequence of multiple transformers under high-dimensional composite constraints explodes exponentially, causing traditional heuristic algorithms to easily fall into local extrema and have slow global convergence performance. Therefore, a parallel dismantling sequence planning method for multiple scrap oil-immersed transformers that considers batch scheduling is proposed. This invention primarily utilizes a dynamic modeling mechanism for integrated process time, a bufferless blocking flow batch scheduling model, and an improved multi-objective evolutionary algorithm. Under the premise of complying with multi-dimensional composite constraints and time domain boundary conditions, it establishes a dynamic operation time model by setting up dismantling priority constraints, tool switching time constraints, and worker fatigue and recovery constraints. Based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, it constructs a bufferless blocking flow recursive completion time model for multiple transformers across multiple workstations. Based on a double-layer chromosome coding structure combined with constraint relationships to maintain crossover and feasible window insertion mutation, it performs multi-objective optimization to achieve Pareto front intelligent sequence planning and decision-making with the joint optimization objectives of minimizing the maximum completion time and maximizing the total profit of the batch dismantling system. This significantly improves batch dismantling efficiency, eliminates process conflicts, increases workstation utilization, and outputs a visualized dismantling solution.

[0007] The technical means employed in this invention are as follows:

[0008] A method for planning the parallel dismantling sequence of multiple scrapped oil-immersed transformers considering batch scheduling includes: S1. Collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data; S2. Based on the collected dismantling data, and in accordance with multidimensional composite constraints, establish an operation time model for the dismantling of waste oil-immersed transformers, and calculate the effective operation time of each dismantling task. S3. For the problem of unbuffered batch scheduling of multiple transformers at multiple serial workstations, based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, a batch decomposition time-schedule matrix of the entry, operation, departure, and blocking times of each transformer at each workstation is established. A unbuffered blocking flow recursive completion time model of multiple transformers at multiple workstations is constructed to calculate the entry and exit times of multiple devices in the pipeline and to calculate the scheduling table of each device in the pipeline. S4. With minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling as the joint optimization objectives, and taking workstation utilization and workstation load balancing as auxiliary optimization objectives, a multi-objective optimization model is constructed. S5. Based on the improved genetic algorithm, under the premise of satisfying all constraints, solve the multi-objective optimization model and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. S6. Convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each station and each transformer, as well as the task content and the iterative convergence curve of the objective function.

[0009] Furthermore, in step S1: The transformer structure data includes the transformer model, dismantling task set, and dismantling task priority relationship; The process parameter data includes disassembly task operation time, tool set, tool changeover time, fatigue attenuation parameters, and recovery parameters; The scheduling resource data includes the number of workstations to be disassembled, workstation-task compatibility constraints, and the number of transformers in the batch.

[0010] Further, step S2 includes: S21. Establish disassembly priority constraints, specifically including: Based on the dismantling constraints of scrap oil-immersed power transformers, a directed graph model of the target transformer is established, transforming the physical structure of the scrap oil-immersed power transformer into a graph theory model. The formula is as follows:

[0011] in, This represents the set of tasks to be disassembled. This represents the total number of scrap power transformers to be dismantled. This represents a set of priority relationships, if we decompose tasks. Must The previous decomposition revealed the existence of directed edges. ; Introducing the adjacency constraint matrix This transforms the graph theory model into a computer-processable mathematical model, where the elements of the adjacency constraint matrix are defined as follows:

[0012] Then any valid decomposition sequence Satisfy topology feasibility constraints:

[0013] In the above formula, Indicates task In the legal disassembly sequence Position index in; Indicates task In the legal disassembly sequence Position index in; Define a valid disassembly sequence. The necessary and sufficient condition for whether the constraint is satisfied is a matrix. All of the above satisfy The task pair, in its valid disassembly sequence All of them meet the task requirements. Appear in the mission Before; S22. Establish tool switching time constraints, specifically including: Introducing matrices This represents the tool switching time during the disassembly process, and the matrix elements are represented as follows:

[0014] Calculation workstation The total switching time for processing task sequence segments is calculated using the following formula:

[0015] In the above formula, Indicates workstation Total switching time for processing task sequence segments; Indicates assignment to workstation The number of tasks contained in the task subset; Indicates when the workstation is from the tool Switch to Tools The time required for tool switching during this process; Indicates workstation Upper The types of disassembly tools required for each task Indicates workstation Upper The types of disassembly tools required for each task; S23. Establish worker fatigue constraints, specifically including: The fatigue decay function is introduced to represent the worker fatigue effect caused by prolonged work, as shown in the following formula:

[0016] In the above formula, Indicates workstation Processing the The cumulative effective working time in front of the transformer; Both represent fatigue attenuation coefficients; This indicates the total working hours of the workers; In a congested flow workshop, the time spent waiting at a workstation is equivalent to the worker's passive rest time. A recovery function is introduced to represent the worker recovery effect caused by congestion and waiting. The recovery function is as follows:

[0017] In the above formula, Indicates the restoration of the function value; Indicates workstation Processing the The blocking time when the transformer is in operation; Indicates the coefficient of restitution; Indicates the recovery rate coefficient; Based on the fatigue decay function and recovery function, the cumulative working time of the worker is updated, and the state update equation is as follows:

[0018] in, Indicates workstation Processing the The total effective working time of workers in front of the transformer; Indicates workstation Processing the The total effective working time of workers in front of the transformer; This indicates the actual operation time of the dismantling task. Indicates the worker's fatigue time. Indicates worker recovery time; Calculate the first The task of the transformer At workstation Actual working time The formula is as follows:

[0019] in, Indicates the first The task of the transformer At workstation The actual working time on the surface; Indicates task Basic dismantling time; Let f be the fatigue decay function, representing the workstation. Processing the The fatigue effect coefficient of workers operating a transformer, among which For workstation Processing the The total effective working time of workers in front of the transformer; S24. Establish workstation-task compatibility constraints, specifically including: set up For the first The task of the transformer Should I be assigned to a workstation? The binary decision variables are defined. Indicates the first The task of the transformer Assigned to workstation , Indicates the first The task of the transformer Not assigned to workstation Then each disassembly task can only be assigned to one workstation, and the following conditions must be met:

[0020] in, , , Indicates the maximum number of tasks; set up For workstation-task compatibility judgment factors, if task Allowed at the workstation Execute, then ,otherwise Then the task-workstation compatibility constraint is:

[0021] in, , , ; This indicates the number of transformers to be dismantled in this batch; Indicates the maximum number of workstations; When no special workstation compatibility restrictions are set Indicates task Can be assigned to any workstation Execution; the workstation number only indicates the available dismantling resource number and does not limit the process sequence of dismantling tasks. The sequence of dismantling tasks is determined by the dismantling priority set. And the task time priority constraints are determined; S25, Calculate the first The transformer in the first The effective working time at a workstation is calculated using the following formula:

[0022] in, Indicates the first The transformer in the first Effective working time at the workstation.

[0023] Further, step S3 includes: S31. Define task scheduling variables and workstation scheduling variables, where... Indicates the first Task in Taiwan transformer The moment the mission begins; Indicates the first Task in Taiwan transformer The time of task completion; Indicates the first The transformer in the first The start time of workstation entry into the workstation formed by the set of tasks corresponding to the workstation; Indicates the first The transformer in the first The completion time of the workstation operation corresponding to the task set; Indicates the first The transformer in the first The actual departure time when the resources of the workstation are released after the task set corresponding to the workstation is completed; Indicates the first The transformer in the first The duration of blocking caused by workstation release, transfer, or failure to meet cache constraints after the completion of the task set corresponding to the workstation; S32. Under the condition of no buffering, the scheduling process must simultaneously satisfy the following constraints: the occupancy constraints of the same workstation before and after the same batch, the flow constraints of adjacent workstations, and the receiving constraints of the downstream workstation, as follows: No. The transformer must be in the previous work station You can only enter your current workstation after you have completed your task and left. ; For the same workstation At most one dismantling task can be executed at any given time. and tasks They were all assigned to workstations If the execution time intervals of the two workstations do not overlap, they will be transferred to the adjacent workstation. Current workstation We must wait for a transformer. Only after completing the task and leaving can one receive the task of dismantling the next transformer; S33. Establish time priority constraints for dismantling tasks, specifically including: For the Priority of arbitrary dismantling of the transformer ,Task The start time must not be earlier than the task. The completion time, that is:

[0024] in, , ; S34. Establish a task scheduling and generation model, specifically including: When the task When all prerequisite tasks have been completed, the task will be completed. Add tasks to the executable task set; for tasks in the executable task set, select a workstation that meets the compatibility conditions for execution based on workstation idle status, tool changeover time, fatigue status, and objective function evaluation results, and determine the task start time. and task completion time ; S35. Establish a workstation time window aggregation model, specifically including: For the The transformer in the first The set of tasks performed at the workstation ,like If not empty, then the first The transformer in the first Workstation entry start time Take the earliest task start time and job completion time in the task set. Take the latest completion time of the task in the task set, that is:

[0025]

[0026] like If the value is empty, the corresponding matrix cell will not generate a workstation occupancy time window; S36. Establish a correction model for the actual departure time, specifically including: Under conditions of no caching or limited caching, when the first The transformer in the first After a workstation completes its corresponding task set, if the constraints on the occupancy of previous and subsequent batches, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations have not yet been satisfied, then the... The transformer continues to occupy the corresponding workstation or scheduling time window until the release conditions are met, and the moment when the release conditions are met is recorded as the actual departure time. ,in:

[0027] S37. Establish a model for calculating blocking duration, specifically including:

[0028] when When, it indicates the first The transformer in the first After the task set corresponding to the workstation is completed, there may be blocking or waiting time for release; when When the time window for that workstation is completed, it indicates that the time can be released directly. S38. Establish a time-space scheduling matrix with transformer serial numbers and workstation two-dimensional indexes, specifically including: The start time of entry into the station, the completion time of the operation, the actual departure time, and the duration of congestion for each transformer at each work station are combined into a batch dismantling time and space scheduling matrix. :

[0029] Through the and By performing task scheduling aggregation and workstation time window correction, a batch scheduling table for multiple scrap oil-immersed transformers in a multi-workstation parallel dismantling scenario is obtained. S39. Based on the set scheduling constraints, establish a recursive completion time model, with the following recursive relationship:

[0030]

[0031] In the above formula, Indicates workstation The cumulative effective working time; Indicates workstation The cumulative effective working time; S310. Determine the recursive boundary conditions as follows:

[0032]

[0033]

[0034] S311. Based on the recursive formula and recursive boundary conditions, starting from the first transformer at the first workstation, the calculation is performed layer by layer to finally obtain the completion time of the last transformer at the last workstation. This time is the completion time of the batch dismantling. The formula is as follows:

[0035] In the above formula, This indicates the total completion time for batch dismantling, that is, the total time required to dismantle all transformers.

[0036] Further, step S4 includes: S41, Total Completion Time Based on Batch Disassembly Establish a target model that minimizes disassembly time. ; S42. Calculate the dismantling revenue, i.e., the revenue from dismantling materials. The formula is as follows:

[0037] S43. Calculate the total dismantling cost. The formula is as follows:

[0038] In the above formula, This represents the cost per unit of time and space. This represents the labor cost per unit of time. S44, Revenue Based on Dismantled Materials Total cost of dismantling Calculate the total profit from dismantling. The formula is as follows:

[0039] S45, Based on total dismantling profit Establish a profit maximization objective model for decomposition, with the following formula:

[0040] S46. Establish an evaluation model for congestion occupancy and workstation utilization, specifically including: Calculate total blocking time The formula is as follows:

[0041] Workstation utilization rate The ratio of the effective working time of each workstation to the total completion time of batch dismantling is determined by the following expression:

[0042] Workstation load balancing index The expression is determined based on the dispersion of the operation time at each workstation, as follows:

[0043] in, The standard deviation of the cumulative effective working time. This represents the average cumulative effective working time. S47. Establish a multi-objective normalization model, specifically including: because , , , and Having different dimensions, for , , , and Perform normalization; assume and These are the upper and lower bounds for the total completion time of batch disassembly, respectively. and These are the upper and lower bounds for the total profit from batch dismantling, respectively. and These are the upper and lower bounds of the total blocking time, respectively. and Let the upper and lower bounds of the workstation utilization rate be the reference upper and lower bounds, respectively.

[0044]

[0045]

[0046]

[0047] S48. Establish a multi-objective optimization function model, specifically including: The weighted objective function is established as follows:

[0048] in, The weighting coefficients are the objective function, and .

[0049] Further, step S5 includes: S51. Generate an initial decomposition sequence that satisfies all constraints, specifically including: S511, Based on the set of priority relations E and adjacency constraint matrix Establish a candidate task set, which contains all decomposition tasks that have no prerequisite tasks or whose prerequisite tasks have been completed. S512. Randomly select a disassembly task from the set and add it to the chromosome sequence; S513, in the set of priority relations E and adjacency constraint matrix Remove the disassembly task from the list and update the constraints of the remaining disassembly tasks. S514. Add newly generated decomposition tasks that have no prerequisite constraints or whose prerequisite tasks have been completed to the candidate task set. S515. Repeat steps S511-S513 until all disassembly tasks are in the sequence. S52. Design a two-layer chromosome coding and decoding method, specifically including: S521. The design includes a two-layer chromosome structure with a first-layer encoding and a second-layer encoding, wherein the first-layer encoding is a sequence of decomposition tasks that satisfy priority constraints; and the second-layer encoding is a workstation allocation scheme for each task. S522. Decode the chromosome into a specific scheduling time. The decoding process is as follows: Based on the first-level sequence, a disassembly order that satisfies the priority relationship is generated; based on the second-level coding, the workstation of each task is determined; based on the recursive completion time model, the processing time of each task is calculated; the completion time of the transformer at each workstation is calculated; the final completion time and the load of each workstation are obtained as the basis for fitness evaluation. S53. Employing an improved genetic algorithm for multi-objective optimization search, specifically including: S531. To avoid generating infeasible solutions during evolution, a constraint relationship is used to maintain crossover. The operation steps are as follows: Generate a binary mask; if the mask bit is 0, select the next unselected task gene from parent generation 1 and add it to the offspring; if the mask bit is 1, select the next unselected task gene from parent generation 2; repeat until the offspring are filled, that is, crossover can only generate offspring if the parent generation satisfies the constraints. S532. For mutation operations, insertion mutation based on feasible windows is adopted. The operation steps are as follows: randomly select a gene in the sequence; in the current sequence, find the latest position of all its predecessor tasks and the earliest position of all its successor tasks; define a feasible window and randomly insert the gene into any position within the window. S533. To ensure that the discovered optimal solution is not lost due to random genetic operations, an elite retention scale is set. In each generation iteration, the individual with the best fitness in the current generation is directly copied to the next generation, merged with the offspring, sorted by fitness, and the best individual is selected to form a new population. S534. Using a tournament selection strategy, individuals are randomly selected each time, and the one with the lowest fitness is selected to enter the next generation. This process is repeated to form a new candidate population. S54, Read the priority relationship set E Adjacency constraint matrix Decompose the weighted coefficients of the objective function Parameters, and simultaneously generate the initial disassembly sequence; S55. Extract the current generation best according to the elite retention strategy; generate the parent candidate set through tournament selection; complete the genetic operation of the genetic algorithm through selection, crossover mutation and constraint repair; evaluate the fitness of offspring; merge elites and offspring, select the best individuals to form the next generation elite; record the current generation best completion time, net profit and workstation utilization rate; use fast non-dominated sorting to screen out the Pareto optimal front to enter the next generation. S56. After reaching the maximum number of iterations, output the optimal disassembly sequence that minimizes disassembly time and maximizes disassembly profit.

[0050] This invention also provides a parallel dismantling sequence planning system for multiple scrap oil-immersed transformers, based on the above-mentioned method for planning parallel dismantling sequences of multiple scrap oil-immersed transformers considering batch scheduling. The system includes: a data input module, a dismantling constraint and operation time modeling module, a bufferless congestion flow batch scheduling modeling module, an optimization objective construction module, a multi-objective optimization solution module, and a visualization output module. The data input module is used to collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data. The dismantling constraint and operation time modeling module is used to establish an operation time model for the dismantling of waste oil-immersed transformers based on the collected dismantling data and in accordance with multi-dimensional composite constraints, and to calculate the effective operation time of each dismantling task. The bufferless blocking flow batch scheduling modeling module addresses the bufferless batch scheduling problem of multiple transformers at multiple serial workstations. Based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, it establishes a batch decomposition time-spacing matrix for the entry, operation, departure, and blocking times of each transformer at each workstation. It also constructs a bufferless blocking flow recursive completion time model for multiple transformers at multiple workstations, calculates the entry and exit times of multiple devices in the pipeline, and calculates the scheduling table for each device in the pipeline. The optimization objective construction module is used to construct a multi-objective optimization model with the joint optimization objective of minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling, and with workstation utilization and workstation load balancing as auxiliary optimization objectives. The multi-objective optimization solution module is used to solve the multi-objective optimization model based on the improved genetic algorithm, under the premise of satisfying all constraints, and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. The visualization output module is used to convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each workstation and each transformer, as well as the task content and the iterative convergence curve of the objective function.

[0051] Compared with the prior art, the present invention has the following advantages: 1. The parallel dismantling sequence planning method for multiple scrap oil-immersed transformers considering batch scheduling provided by this invention is based on dynamic modeling of integrated process time, a bufferless blocking flow recursive completion time model and an improved multi-objective evolutionary algorithm. It achieves a significant reduction in the total completion time of batch dismantling compared to dismantling one transformer at a time and a significant reduction in the average dismantling time per transformer, thus significantly improving the efficiency of batch dismantling.

[0052] 2. The parallel dismantling sequence planning method for multiple scrap oil-immersed transformers considering batch scheduling provided by this invention, through unified mathematical modeling with dismantling priority constraints, ensures that all dismantling operations meet the dismantling task priority constraints, fundamentally eliminating process conflicts.

[0053] 3. The parallel dismantling sequence planning method for multiple scrap oil-immersed transformers considering batch scheduling provided by this invention is based on the multi-objective optimization solution strategy of the improved genetic algorithm. It utilizes a double-layer chromosome coding structure combined with constraint-maintaining crossover and feasible window insertion mutation mechanism to keep the utilization rate of each workstation within a reasonable range, which significantly improves the average workstation utilization rate compared with dismantling one transformer at a time.

[0054] 4. The parallel dismantling sequence planning method for multiple scrap oil-immersed transformers considering batch scheduling provided by this invention, with the help of two-dimensional scheduling matrix and batch dismantling Gantt chart conversion technology, realizes the intuitive presentation of dismantling progress and dismantling sequence of each dismantling task, which facilitates on-site operation guidance and reduces manual dependence.

[0055] 5. The parallel dismantling sequence planning method for multiple scrap oil-immersed transformers provided by this invention, which considers batch scheduling, has a flexible adjustment mechanism based on batch size, number of workstations, and dismantling task constraints, and achieves wide adaptability to the batch dismantling needs of oil-immersed power transformers of different models and recycling scenarios.

[0056] In summary, the technical solution of this invention solves the problems of low batch disassembly efficiency, process conflicts and workstation load imbalance, lack of physical mechanism for unbuffered blocking flow, insufficient adaptability of traditional algorithms to multi-constraint batch optimization, and low degree of visualization in the prior art.

[0057] Based on the above reasons, this invention can be widely promoted in fields such as large-scale recycling production lines for decommissioned oil-immersed power transformers, intensive disposal centers for waste power materials, and digital twin scenarios for intelligent dismantling production lines driven by industrial cyber-physical systems. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This is a flowchart of the method of the present invention.

[0060] Figure 2 A directed graph provided for an embodiment of the present invention.

[0061] Figure 3 A schematic diagram of the dismantling site provided for an embodiment of the present invention.

[0062] Figure 4 A flowchart for generating the disassembly sequence is provided for this invention.

[0063] Figure 5 This is a block diagram of the system structure of the present invention.

[0064] Figure 6 This is a schematic diagram of the bufferless blocking stream mechanism provided by the present invention.

[0065] Figure 7 The diagram shows the optimization effect of disassembly time, disassembly profit, workstation utilization, and blockage time provided in this embodiment.

[0066] Figure 8 This is a Gantt chart of the optimal disassembly sequence provided in the embodiments of the present invention.

[0067] Figure 9 The workstation load diagram provided for an embodiment of the present invention.

[0068] Figure 10 This is a performance comparison chart of different batch dismantling and scheduling methods provided in the embodiments of the present invention. Detailed Implementation

[0069] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0070] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product or device.

[0071] like Figure 1 As shown, this invention provides a method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling, including: S1. Collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data; S2. Based on the collected dismantling data, and in accordance with multidimensional composite constraints, establish an operation time model for the dismantling of waste oil-immersed transformers, and calculate the effective operation time of each dismantling task. S3. For the problem of unbuffered batch scheduling of multiple transformers at multiple serial workstations, based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, a batch decomposition time-schedule matrix of the entry, operation, departure, and blocking times of each transformer at each workstation is established. A unbuffered blocking flow recursive completion time model of multiple transformers at multiple workstations is constructed to calculate the entry and exit times of multiple devices in the pipeline and to calculate the scheduling table of each device in the pipeline. S4. With minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling as the joint optimization objectives, and taking workstation utilization and workstation load balancing as auxiliary optimization objectives, a multi-objective optimization model is constructed. S5. Based on the improved genetic algorithm, under the premise of satisfying all constraints, solve the multi-objective optimization model and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. S6. Convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each station and each transformer, as well as the task content and the iterative convergence curve of the objective function.

[0072] In a specific implementation, as a preferred embodiment of the present invention, in step S1: The transformer structure data includes the transformer model, dismantling task set, and dismantling task priority relationship; The process parameter data includes disassembly task operation time, tool set, tool changeover time, fatigue attenuation parameters, and recovery parameters; The scheduling resource data includes the number of workstations to be disassembled, workstation-task compatibility constraints, and the number of transformers in the batch.

[0073] In a specific implementation, as a preferred embodiment of the present invention, step S2 includes: S21. Transformer dismantling consists of multiple tasks, and the dismantling process must follow the priority constraints between these tasks. Therefore, it is necessary to establish dismantling priority order constraints, specifically including: Based on the dismantling constraints of scrap oil-immersed power transformers, such as Figure 2 As shown, a directed graph model of the target transformer is established, transforming the physical structure of the scrap oil-immersed power transformer into a graph theory model. The formula is as follows:

[0074] in, This represents the set of tasks to be disassembled. This represents the total number of scrap power transformers to be dismantled. This represents a set of priority relationships, if we decompose tasks. Must The previous decomposition revealed the existence of directed edges. ; Introducing the adjacency constraint matrix This transforms the graph theory model into a computer-processable mathematical model, where the elements of the adjacency constraint matrix are defined as follows:

[0075] Then any valid decomposition sequence Satisfy topology feasibility constraints:

[0076] In the above formula, Indicates task In the legal disassembly sequence Position index in; Indicates task In the legal disassembly sequence Position index in; Define a valid disassembly sequence. The necessary and sufficient condition for whether the constraint is satisfied is a matrix. All of the above satisfy The task pair, in its valid disassembly sequence All of them meet the task requirements. Appear in the mission Before; S22. Transformer dismantling involves various tools such as oil pumps and lifting equipment. Different tools are required for different dismantling tasks. For large power equipment like transformers, the time spent switching tools should not be ignored. Therefore, it is necessary to establish tool switching time constraints, specifically including: Introducing matrices This represents the tool switching time during the disassembly process, and the matrix elements are represented as follows:

[0077] Calculation workstation The total switching time for processing task sequence segments is calculated using the following formula:

[0078] In the above formula, Indicates workstation Total switching time for processing task sequence segments; Indicates assignment to workstation The number of tasks contained in the task subset; Indicates when the workstation is from the tool Switch to Tools The time required for tool switching during this process; Indicates workstation Upper The types of disassembly tools required for each task Indicates workstation Upper The types of disassembly tools required for each task; S23. Prolonged continuous dismantling operations can lead to worker fatigue and reduce work efficiency. Therefore, it is necessary to establish worker fatigue constraints, which specifically include: The fatigue decay function is introduced to represent the worker fatigue effect caused by prolonged work, as shown in the following formula:

[0079] In the above formula, Indicates workstation Processing the The cumulative effective working time in front of the transformer; Both represent fatigue attenuation coefficients; This indicates the total working hours of the workers; In a congested flow workshop, the time spent waiting at a workstation is equivalent to the worker's passive rest time. A recovery function is introduced to represent the worker recovery effect caused by congestion and waiting. The recovery function is as follows:

[0080] In the above formula, Indicates the restoration of the function value; Indicates workstation Processing the The blocking time when the transformer is in operation; Indicates the coefficient of restitution; Indicates the recovery rate coefficient; Based on the fatigue decay function and recovery function, the cumulative working time of the worker is updated, and the state update equation is as follows:

[0081] in, Indicates workstation Processing the The total effective working time of workers in front of the transformer; Indicates workstation Processing the The total effective working time of workers in front of the transformer; This indicates the actual operation time of the dismantling task. Indicates the worker's fatigue time. Indicates worker recovery time; Calculate the first The task of the transformer At workstation Actual working time The formula is as follows:

[0082] in, Indicates the first The task of the transformer At workstation The actual working time on the surface; Indicates task Basic dismantling time; Let f be the fatigue decay function, representing the workstation. Processing the The fatigue effect coefficient of workers operating a transformer, among which For workstation Processing the The total effective working time of workers in front of the transformer; S24. Establish workstation-task compatibility constraints, specifically including: set up For the first The task of the transformer Should I be assigned to a workstation? The binary decision variables are defined. Indicates the first The task of the transformer Assigned to workstation , Indicates the first The task of the transformer Not assigned to workstation Then each disassembly task can only be assigned to one workstation, and the following conditions must be met:

[0083] in, , , Indicates the maximum number of tasks; set up For workstation-task compatibility judgment factors, if task Allowed at the workstation Execute, then ,otherwise Then the task-workstation compatibility constraint is:

[0084] in, , , ; This indicates the number of transformers to be dismantled in this batch; Indicates the maximum number of workstations; When no special workstation compatibility restrictions are set Indicates task Can be assigned to any workstation Execution; the workstation number only indicates the available dismantling resource number and does not limit the process sequence of dismantling tasks. The sequence of dismantling tasks is determined by the dismantling priority set. And the task time priority constraints are determined; S25, Calculate the first The transformer in the first The effective working time at a workstation is calculated using the following formula:

[0085] in, Indicates the first The transformer in the first Effective working time at the workstation.

[0086] This embodiment uses a certain type of oil-immersed transformer as the dismantling object, and optimizes the dismantling of 5 units in batches at 4 workstations. The dismantling site is as follows: Figure 3 As shown in Table 1, the basic data of this type of oil-immersed transformer are as follows.

[0087] Table 1 Basic Data of Oil-Immersed Transformers

[0088] In a specific implementation, as a preferred embodiment of the present invention, step S3 includes: S31. Define task scheduling variables and workstation scheduling variables, where... Indicates the first Task in Taiwan transformer The moment the mission begins; Indicates the first Task in Taiwan transformer The time of task completion; Indicates the first The transformer in the first The start time of workstation entry into the workstation formed by the set of tasks corresponding to the workstation; Indicates the first The transformer in the first The completion time of the workstation operation corresponding to the task set; Indicates the first The transformer in the first The actual departure time when the resources of the workstation are released after the task set corresponding to the workstation is completed; Indicates the first The transformer in the first The duration of blocking caused by workstation release, transfer, or failure to meet cache constraints after the completion of the task set corresponding to the workstation; The dismantling workshop for S32 oil-immersed transformers primarily features a serial workstation structure in its physical layout. However, transformers are large, and on-site working space is limited, making it impossible to set up intermediate buffer zones between workstations. When a downstream workstation has not yet completed its work, an upstream workstation, even if it has completed its current task, cannot transfer the workpiece, thus forcing it to occupy workstation resources and causing congestion. Therefore, under the condition of no buffer, it is necessary to set constraints that the scheduling process must simultaneously satisfy: batch occupancy constraints of the same workstation, flow constraints of adjacent workstations, and receiving constraints of the downstream workstation, as follows: No. The transformer must be in the previous work station You can only enter your current workstation after you have completed your task and left. ; For the same workstation At most one dismantling task can be executed at any given time. and tasks They were all assigned to workstations If the execution time intervals of the two workstations do not overlap, they will be transferred to the adjacent workstation. Current workstation We must wait for a transformer. Only after completing the task and leaving can one receive the task of dismantling the next transformer; S33. Establish time priority constraints for dismantling tasks, specifically including: For the Priority of arbitrary dismantling of the transformer ,Task The start time must not be earlier than the task. The completion time, that is:

[0089] in, , ; S34. Establish a task scheduling and generation model, specifically including: When the task When all prerequisite tasks have been completed, the task will be completed. Add tasks to the executable task set; for tasks in the executable task set, select a workstation that meets the compatibility conditions for execution based on workstation idle status, tool changeover time, fatigue status, and objective function evaluation results, and determine the task start time. and task completion time ; S35. Establish a workstation time window aggregation model, specifically including: For the The transformer in the first The set of tasks performed at the workstation ,like If not empty, then the first The transformer in the first Workstation entry start time Take the earliest task start time and job completion time in the task set. Take the latest completion time of the task in the task set, that is:

[0090]

[0091] like If the value is empty, the corresponding matrix cell will not generate a workstation occupancy time window; S36. Establish a correction model for the actual departure time, specifically including: Under conditions of no caching or limited caching, when the first The transformer in the first After a workstation completes its corresponding task set, if the constraints on the occupancy of previous and subsequent batches, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations have not yet been satisfied, then the... The transformer continues to occupy the corresponding workstation or scheduling time window until the release conditions are met, and the moment when the release conditions are met is recorded as the actual departure time. ,in:

[0092] S37. Establish a model for calculating blocking duration, specifically including:

[0093] when When, it indicates the first The transformer in the first After the task set corresponding to the workstation is completed, there may be blocking or waiting time for release; when When the time window for that workstation is completed, it indicates that the time can be released directly. S38. Establish a time-space scheduling matrix with transformer serial numbers and workstation two-dimensional indexes, specifically including: The start time of entry into the station, the completion time of the operation, the actual departure time, and the duration of congestion for each transformer at each work station are combined into a batch dismantling time and space scheduling matrix. :

[0094] Through the and By performing task scheduling aggregation and workstation time window correction, a batch scheduling table for multiple scrap oil-immersed transformers in a multi-workstation parallel dismantling scenario is obtained. S39. Based on the set scheduling constraints, establish a recursive completion time model, with the following recursive relationship:

[0095]

[0096] In the above formula, Indicates workstation The cumulative effective working time; Indicates workstation The cumulative effective working time; S310. Determine the recursive boundary conditions as follows:

[0097]

[0098]

[0099] S311. Based on the recursive formula and recursive boundary conditions, starting from the first transformer at the first workstation, the calculation is performed layer by layer to finally obtain the completion time of the last transformer at the last workstation. This time is the completion time of the batch dismantling. The formula is as follows:

[0100] In the above formula, This indicates the total completion time for batch dismantling, that is, the total time required to dismantle all transformers.

[0101] In a specific implementation, as a preferred embodiment of the present invention, step S4 includes: S41, Total Completion Time Based on Batch Disassembly Establish a target model that minimizes disassembly time. ; S42. Calculate the dismantling revenue, i.e., the revenue from dismantling materials. The formula is as follows:

[0102] S43. Calculate the total dismantling cost. The formula is as follows:

[0103] In the above formula, This represents the cost per unit of time and space. This represents the labor cost per unit of time. S44, Revenue Based on Dismantled Materials Total cost of dismantling Calculate the total profit from dismantling. The formula is as follows:

[0104] S45, Based on total dismantling profit Establish a profit maximization objective model for decomposition, with the following formula:

[0105] S46. Establish an evaluation model for congestion occupancy and workstation utilization, specifically including: Calculate total blocking time The formula is as follows:

[0106] Workstation utilization rate The ratio of the effective working time of each workstation to the total completion time of batch dismantling is determined by the following expression:

[0107] Workstation load balancing index The expression is determined based on the dispersion of the operation time at each workstation, as follows:

[0108] in, The standard deviation of the cumulative effective working time. This represents the average cumulative effective working time. S47. Establish a multi-objective normalization model, specifically including: because , , , and Having different dimensions, for , , , and Perform normalization; assume and These are the upper and lower bounds for the total completion time of batch disassembly, respectively. and These are the upper and lower bounds for the total profit from batch dismantling, respectively. and These are the upper and lower bounds of the total blocking time, respectively. and Let the upper and lower bounds of the workstation utilization rate be the reference upper and lower bounds, respectively.

[0109]

[0110]

[0111]

[0112] S48. Establish a multi-objective optimization function model, specifically including: The weighted objective function is established as follows:

[0113] in, The weighting coefficients are the objective function, and .

[0114] In a specific implementation, as a preferred embodiment of the present invention, step S5 includes: S51. Generate an initial decomposition sequence that satisfies all constraints, such as... Figure 4 As shown, it specifically includes: S511, Based on the set of priority relations E and adjacency constraint matrix Establish a candidate task set, which contains all decomposition tasks that have no prerequisite tasks or whose prerequisite tasks have been completed. S512. Randomly select a disassembly task from the set and add it to the chromosome sequence; S513, in the set of priority relations E and adjacency constraint matrix Remove the disassembly task from the list and update the constraints of the remaining disassembly tasks. S514. Add newly generated decomposition tasks that have no prerequisite constraints or whose prerequisite tasks have been completed to the candidate task set. S515. Repeat steps S511-S513 until all disassembly tasks are in the sequence. S52. Design a two-layer chromosome coding and decoding method, specifically including: S521. Since two types of decision variables are involved simultaneously, namely the priority order of component disassembly and the allocation relationship of tasks in each workstation, a two-layer chromosome structure including a first-layer coding and a second-layer coding is designed. The first-layer coding is the disassembly task sequence that satisfies the priority constraint; the second-layer coding is the workstation allocation scheme corresponding to each task. S522. Since the quality of individuals in a genetic algorithm depends on their actual performance in the model, chromosomes need to be decoded into specific scheduling times. The decoding process is as follows: Based on the first-level sequence, a disassembly order that satisfies the priority relationship is generated; based on the second-level coding, the workstation of each task is determined; based on the recursive completion time model, the processing time of each task is calculated; the completion time of the transformer at each workstation is calculated; the final completion time and the load of each workstation are obtained as the basis for fitness evaluation. S53. After completing the above encoding and decoding mechanism design, since the problem involves a mixture of discrete sequence planning and continuous resource allocation, the genetic algorithm has good robustness in handling such discrete combinatorial optimization problems. Furthermore, to improve the algorithm's adaptability to complex constraint problems, an improved genetic algorithm is used for multi-objective optimization search, specifically including: S531. To avoid generating infeasible solutions during evolution, a constraint relationship is used to maintain crossover. The operation steps are as follows: Generate a binary mask; if the mask bit is 0, select the next unselected task gene from parent generation 1 and add it to the offspring; if the mask bit is 1, select the next unselected task gene from parent generation 2; repeat until the offspring are filled, that is, crossover can only generate offspring if the parent generation satisfies the constraints. S532. For mutation operations, insertion mutation based on feasible windows is adopted. The operation steps are as follows: randomly select a gene in the sequence; in the current sequence, find the latest position of all its predecessor tasks and the earliest position of all its successor tasks; define a feasible window and randomly insert the gene into any position within the window, thereby ensuring that the mutated individual is still topologically feasible. S533. To ensure that the discovered optimal solution is not lost due to random genetic operations, an elite retention scale is set. In each generation iteration, the individual with the best fitness in the current generation is directly copied to the next generation, merged with the offspring, sorted by fitness, and the best individual is selected to form a new population. S534. Using a tournament selection strategy, individuals are randomly selected each time, and the one with the lowest fitness is selected to enter the next generation. This process is repeated to form a new candidate population. S54, Read the priority relationship set E Adjacency constraint matrix Decompose the weighted coefficients of the objective function Parameters, and simultaneously generate the initial disassembly sequence; S55. Extract the current generation best according to the elite retention strategy; generate the parent candidate set through tournament selection; complete the genetic operation of the genetic algorithm through selection, crossover mutation and constraint repair; evaluate the fitness of offspring; merge elites and offspring, select the best individuals to form the next generation elite; record the current generation best completion time, net profit and workstation utilization rate; use fast non-dominated sorting to screen out the Pareto optimal front to enter the next generation. S56. After reaching the maximum number of iterations, output the optimal disassembly sequence that minimizes disassembly time and maximizes disassembly profit.

[0115] This invention also provides a parallel dismantling sequence planning system for multiple scrap oil-immersed transformers, based on the above-mentioned method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling. Figure 5 As shown, it includes: a data input module, a constraint decomposition and job time modeling module, a bufferless blocking flow batch scheduling modeling module, an optimization objective construction module, a multi-objective optimization solution module, and a visualization output module, wherein: The data input module is used to collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data. The dismantling constraint and operation time modeling module is used to establish an operation time model for the dismantling of waste oil-immersed transformers based on the collected dismantling data and in accordance with multi-dimensional composite constraints, and to calculate the effective operation time of each dismantling task. The bufferless blocking flow batch scheduling modeling module addresses the bufferless batch scheduling problem of multiple transformers at multiple serial workstations. Based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, it establishes a batch decomposition time-spacing matrix for the entry, operation, departure, and blocking times of each transformer at each workstation. It also constructs a bufferless blocking flow recursive completion time model for multiple transformers at multiple workstations, calculates the entry and exit times of multiple devices in the pipeline, and calculates the scheduling table for each device in the pipeline. The optimization objective construction module is used to construct a multi-objective optimization model with the joint optimization objective of minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling, and with workstation utilization and workstation load balancing as auxiliary optimization objectives. The multi-objective optimization solution module is used to solve the multi-objective optimization model based on the improved genetic algorithm, under the premise of satisfying all constraints, and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. The visualization output module is used to convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each workstation and each transformer, as well as the task content and the iterative convergence curve of the objective function.

[0116] like Figure 7The diagram illustrates the iterative convergence process of the objective function when optimizing the batch dismantling and scheduling scheme for multiple scrap oil-immersed transformers in this embodiment. As the number of iterations increases, the total batch completion time and total congestion time gradually decrease, while the total batch profit and average workstation utilization gradually increase. This indicates that the population continuously eliminates individuals with poor scheduling performance during the evolutionary process, retaining scheduling schemes with superior overall performance in terms of time efficiency, economic benefits, congestion control, and workstation utilization. In the later stages of iteration, the curves of each indicator gradually stabilize, and the population average gradually approaches the current optimal value, indicating that the improved genetic algorithm has good convergence stability. This demonstrates that the present invention can effectively improve dismantling efficiency and dismantling profit in this embodiment.

[0117] like Figure 8 The diagram shows the optimal dismantling sequence Gantt chart output in this embodiment. Each rectangle represents a dismantling task, and the number within the rectangle indicates the task number. Different shades of gray distinguish different transformers to be dismantled. Different line types of the rectangle borders differentiate the types of tools required for each task. The narrow black blocks represent the tool switching time between adjacent tasks at the same workstation. Therefore, this embodiment visualizes the dismantling scheduling process of multiple transformers at different workstations, intuitively reflecting the parallel dismantling process of multiple transformers under multi-workstation conditions, as well as the scheduling relationship between different transformer tasks.

[0118] like Figure 9 The diagram shown is the workstation load diagram corresponding to the optimized scheduling scheme in this embodiment. This result is used to evaluate the workstation utilization rate, congestion level, and load balance of the optimized scheduling scheme, indicating that the workstation load balance in this embodiment is at a high level.

[0119] like Figure 10 The figure shows a comparison of different batch dismantling and scheduling methods in this embodiment. The figure compares the performance differences of the serial dismantling method, the ideal scheduling method ignoring congestion, the method of this invention, and the congestion correction method in terms of total batch completion time, workstation utilization, and total batch profit.

[0120] The sequential dismantling method involves completely dismantling one transformer before processing the next, resulting in a long overall batch completion time and low workstation utilization. While the ideal scheduling method, which ignores congestion, can shorten the overall completion time through parallel multi-workstation operations, it does not fully consider practical constraints such as workstation release, task connection, and congestion occupancy, thus the results have a certain degree of idealization.

[0121] The blocking correction method modifies the executability of the ideal scheduling scheme, but since it does not consider the coupling relationship between task sequence, workstation allocation and blocking occupation in the optimization process, there is still room for improvement in overall performance.

[0122] In comparison, the total batch completion time of the method of this invention is lower than that of other comparative methods, the workstation utilization rate is higher than that of other methods, and the total batch profit is also maintained at a high level. This result shows that, under the premise of ensuring the feasibility of dismantling task constraints, this invention can improve the batch dismantling efficiency and workstation utilization of multiple transformers operating at multiple workstations, reduce ineffective waiting and resource idleness, and improve the economics of batch dismantling operations.

[0123] The embodiments of the present invention are described simply because they correspond to those in the embodiments above. For any similarities, please refer to the descriptions in the embodiments above, which will not be elaborated here.

[0124] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling, characterized in that, include: S1. Collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data; S2. Based on the collected dismantling data, and in accordance with multidimensional composite constraints, establish an operation time model for the dismantling of waste oil-immersed transformers, and calculate the effective operation time of each dismantling task. S3. For the problem of unbuffered batch scheduling of multiple transformers at multiple serial workstations, based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, a batch decomposition time-schedule matrix of the entry, operation, departure, and blocking times of each transformer at each workstation is established. A unbuffered blocking flow recursive completion time model of multiple transformers at multiple workstations is constructed to calculate the entry and exit times of multiple devices in the pipeline and to calculate the scheduling table of each device in the pipeline. S4. With minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling as the joint optimization objectives, and taking workstation utilization and workstation load balancing as auxiliary optimization objectives, a multi-objective optimization model is constructed. S5. Based on the improved genetic algorithm, under the premise of satisfying all constraints, solve the multi-objective optimization model and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. S6. Convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each station and each transformer, as well as the task content and the iterative convergence curve of the objective function.

2. The method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling as described in claim 1, characterized in that, In step S1: The transformer structure data includes the transformer model, dismantling task set, and dismantling task priority relationship; The process parameter data includes disassembly task operation time, tool set, tool changeover time, fatigue attenuation parameters, and recovery parameters; The scheduling resource data includes the number of workstations to be disassembled, workstation-task compatibility constraints, and the number of transformers in the batch.

3. The method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling as described in claim 1, characterized in that, Step S2 includes: S21. Establish disassembly priority constraints, specifically including: Based on the dismantling constraints of scrap oil-immersed power transformers, a directed graph model of the target transformer is established, transforming the physical structure of the scrap oil-immersed power transformer into a graph theory model. The formula is as follows: in, This represents the set of tasks to be disassembled. This represents the total number of scrap power transformers to be dismantled. This represents a set of priority relationships, if we decompose tasks. Must The previous decomposition revealed the existence of directed edges. ; Introducing the adjacency constraint matrix This transforms the graph theory model into a computer-processable mathematical model, where the elements of the adjacency constraint matrix are defined as follows: Then any valid decomposition sequence Satisfy topology feasibility constraints: In the above formula, Indicates task In the legal disassembly sequence Position index in; Indicates task In the legal disassembly sequence Position index in; Define a valid disassembly sequence. The necessary and sufficient condition for whether the constraint is satisfied is a matrix. All of the above satisfy The task pair, in its valid disassembly sequence All of them meet the task requirements. Appear in the mission Before; S22. Establish tool switching time constraints, specifically including: Introducing matrices This represents the tool switching time during the disassembly process, and the matrix elements are represented as follows: Calculation workstation The total switching time for processing task sequence segments is calculated using the following formula: In the above formula, Indicates workstation Total switching time for processing task sequence segments; Indicates assignment to workstation The number of tasks contained in the task subset; Indicates when the workstation is from the tool Switch to Tools The time required for tool switching during this process; Indicates workstation Upper The types of disassembly tools required for each task Indicates workstation Upper The types of disassembly tools required for each task; S23. Establish worker fatigue constraints, specifically including: The fatigue decay function is introduced to represent the worker fatigue effect caused by prolonged work, as shown in the following formula: In the above formula, Indicates workstation Processing the The cumulative effective working time in front of the transformer; Both represent fatigue attenuation coefficients; This indicates the total working hours of the workers; In a congested flow workshop, the time spent waiting at a workstation is equivalent to the worker's passive rest time. A recovery function is introduced to represent the worker recovery effect caused by congestion and waiting. The recovery function is as follows: In the above formula, Indicates the restoration of the function value; Indicates workstation Processing the The blocking time when the transformer is in operation; Indicates the coefficient of restitution; Indicates the recovery rate coefficient; Based on the fatigue decay function and recovery function, the cumulative working time of the worker is updated, and the state update equation is as follows: in, Indicates workstation Processing the The total effective working time of workers in front of the transformer; Indicates workstation Processing the The total effective working time of workers in front of the transformer; This indicates the actual operation time of the dismantling task. Indicates the worker's fatigue time. Indicates worker recovery time; Calculate the first The task of the transformer At workstation Actual working time The formula is as follows: in, Indicates the first The task of the transformer At workstation The actual working time on the surface; Indicates task Basic dismantling time; Let f be the fatigue decay function, representing the workstation. Processing the The fatigue effect coefficient of workers operating a transformer, among which For workstation Processing the The total effective working time of workers in front of the transformer; S24. Establish workstation-task compatibility constraints, specifically including: set up For the first The task of the transformer Should I be assigned to a workstation? The binary decision variables are defined. Indicates the first The task of the transformer Assigned to workstation , Indicates the first The task of the transformer Not assigned to workstation Then each disassembly task can only be assigned to one workstation, and the following conditions must be met: in, , , Indicates the maximum number of tasks; set up For workstation-task compatibility judgment factors, if task Allowed at the workstation Execute, then ,otherwise Then the task-workstation compatibility constraint is: in, , , ; This indicates the number of transformers to be dismantled in this batch; Indicates the maximum number of workstations; When no special workstation compatibility restrictions are set Indicates task Can be assigned to any workstation Execution; the workstation number only indicates the available dismantling resource number and does not limit the process sequence of dismantling tasks. The sequence of dismantling tasks is determined by the dismantling priority set. And the task time priority constraints are determined; S25, Calculate the first The transformer in the first The effective working time at a workstation is calculated using the following formula: in, Indicates the first The transformer in the first Effective working time at the workstation.

4. The method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling as described in claim 1, characterized in that, Step S3 includes: S31. Define task scheduling variables and workstation scheduling variables, where... Indicates the first Task in Taiwan transformer The moment the mission begins; Indicates the first Task in Taiwan transformer The time of task completion; Indicates the first The transformer in the first The start time of workstation entry into the workstation formed by the set of tasks corresponding to the workstation; Indicates the first The transformer in the first The completion time of the workstation operation corresponding to the task set; Indicates the first The transformer in the first The actual departure time when the resources of the workstation are released after the task set corresponding to the workstation is completed; Indicates the first The transformer in the first The duration of blocking caused by workstation release, transfer, or failure to meet cache constraints after the completion of the task set corresponding to the workstation; S32. Under the condition of no buffering, the scheduling process must simultaneously satisfy the following constraints: the occupancy constraints of the same workstation before and after the same batch, the flow constraints of adjacent workstations, and the receiving constraints of the downstream workstation, as follows: No. The transformer must be in the previous work station You can only enter your current workstation after you have completed your task and left. ; For the same workstation At most one dismantling task can be executed at any given time. and tasks They were all assigned to workstations If the execution time intervals of the two workstations do not overlap, they will be transferred to the adjacent workstation. Current workstation We must wait for a transformer. Only after completing the task and leaving can one receive the task of dismantling the next transformer; S33. Establish time priority constraints for dismantling tasks, specifically including: For the Priority of arbitrary dismantling of the transformer ,Task The start time must not be earlier than the task. The completion time, that is: in, , ; S34. Establish a task scheduling and generation model, specifically including: When the task When all prerequisite tasks have been completed, the task will be completed. Add tasks to the executable task set; for tasks in the executable task set, select a workstation that meets the compatibility conditions for execution based on workstation idle status, tool changeover time, fatigue status, and objective function evaluation results, and determine the task start time. and task completion time ; S35. Establish a workstation time window aggregation model, specifically including: For the The transformer in the first The set of tasks performed at the workstation ,like If not empty, then the first The transformer in the first Workstation entry start time Take the earliest task start time and job completion time in the task set. Take the latest completion time of the task in the task set, that is: like If the value is empty, the corresponding matrix cell will not generate a workstation occupancy time window; S36. Establish a correction model for the actual departure time, specifically including: Under conditions of no caching or limited caching, when the first The transformer in the first After a workstation completes its corresponding task set, if the constraints on the occupancy of previous and subsequent batches, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations have not yet been satisfied, then the... The transformer continues to occupy the corresponding workstation or scheduling time window until the release conditions are met, and the moment when the release conditions are met is recorded as the actual departure time. ,in: S37. Establish a model for calculating blocking duration, specifically including: when When, it indicates the first The transformer in the first After the task set corresponding to the workstation is completed, there may be blocking or waiting time for release; when When the time window for that workstation is completed, it indicates that the time can be released directly. S38. Establish a time-space scheduling matrix with transformer serial numbers and workstation two-dimensional indexes, specifically including: The start time of entry into the station, the completion time of the operation, the actual departure time, and the duration of congestion for each transformer at each work station are combined into a batch dismantling time and space scheduling matrix. : Through the and By performing task scheduling aggregation and workstation time window correction, a batch scheduling table for multiple scrap oil-immersed transformers in a multi-workstation parallel dismantling scenario is obtained. S39. Based on the set scheduling constraints, establish a recursive completion time model, with the following recursive relationship: In the above formula, Indicates workstation The cumulative effective working time; Indicates workstation The cumulative effective working time; S310. Determine the recursive boundary conditions as follows: S311. Based on the recursive formula and recursive boundary conditions, starting from the first transformer at the first workstation, the calculation is performed layer by layer to finally obtain the completion time of the last transformer at the last workstation. This time is the completion time of the batch dismantling. The formula is as follows: In the above formula, This indicates the total completion time for batch dismantling, that is, the total time required to dismantle all transformers.

5. The method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling as described in claim 1, characterized in that, Step S4 includes: S41, Total Completion Time Based on Batch Disassembly Establish a target model that minimizes disassembly time. ; S42. Calculate the dismantling revenue, i.e., the revenue from dismantling materials. The formula is as follows: S43. Calculate the total dismantling cost. The formula is as follows: In the above formula, This represents the cost per unit of time and space. This represents the labor cost per unit of time. S44, Revenue Based on Dismantled Materials Total cost of dismantling Calculate the total profit from dismantling. The formula is as follows: S45, Based on total dismantling profit Establish a profit maximization objective model for decomposition, with the following formula: S46. Establish an evaluation model for congestion occupancy and workstation utilization, specifically including: Calculate total blocking time The formula is as follows: Workstation utilization rate The ratio of the effective working time of each workstation to the total completion time of batch dismantling is determined by the following expression: Workstation load balancing index The expression is determined based on the dispersion of the operation time at each workstation, as follows: in, The standard deviation of the cumulative effective working time. This represents the average cumulative effective working time. S47. Establish a multi-objective normalization model, specifically including: because , , , and Having different dimensions, for , , , and Perform normalization; assume and These are the upper and lower bounds for the total completion time of batch disassembly, respectively. and These are the upper and lower bounds for the total profit from batch dismantling, respectively. and These are the upper and lower bounds of the total blocking time, respectively. and Let the upper and lower bounds of the workstation utilization rate be the reference upper and lower bounds, respectively. S48. Establish a multi-objective optimization function model, specifically including: The weighted objective function is established as follows: in, The weighting coefficients are the objective function, and .

6. The method for planning the parallel dismantling sequence of multiple scrap oil-immersed transformers considering batch scheduling as described in claim 1, characterized in that, Step S5 includes: S51. Generate an initial decomposition sequence that satisfies all constraints, specifically including: S511, Based on the set of priority relations E and adjacency constraint matrix Establish a candidate task set, which contains all decomposition tasks that have no prerequisite tasks or whose prerequisite tasks have been completed. S512. Randomly select a disassembly task from the set and add it to the chromosome sequence; S513, in the set of priority relations E and adjacency constraint matrix Remove the disassembly task from the list and update the constraints of the remaining disassembly tasks. S514. Add newly generated decomposition tasks that have no prerequisite constraints or whose prerequisite tasks have been completed to the candidate task set. S515. Repeat steps S511-S513 until all disassembly tasks are in the sequence. S52. Design a two-layer chromosome coding and decoding method, specifically including: S521. The design includes a two-layer chromosome structure with a first-layer encoding and a second-layer encoding, wherein the first-layer encoding is a sequence of decomposition tasks that satisfy priority constraints; and the second-layer encoding is a workstation allocation scheme for each task. S522. Decode the chromosome into a specific scheduling time. The decoding process is as follows: Based on the first-level sequence, a disassembly order that satisfies the priority relationship is generated; based on the second-level coding, the workstation of each task is determined; based on the recursive completion time model, the processing time of each task is calculated; the completion time of the transformer at each workstation is calculated; the final completion time and the load of each workstation are obtained as the basis for fitness evaluation. S53. Employing an improved genetic algorithm for multi-objective optimization search, specifically including: S531. To avoid generating infeasible solutions during evolution, a constraint relationship is used to maintain crossover. The operation steps are as follows: Generate a binary mask; if the mask bit is 0, select the next unselected task gene from parent generation 1 and add it to the offspring; if the mask bit is 1, select the next unselected task gene from parent generation 2; repeat until the offspring are filled, that is, crossover can only generate offspring if the parent generation satisfies the constraints. S532. For mutation operations, insertion mutation based on feasible windows is adopted. The operation steps are as follows: randomly select a gene in the sequence; in the current sequence, find the latest position of all its predecessor tasks and the earliest position of all its successor tasks; define a feasible window and randomly insert the gene into any position within the window. S533. To ensure that the discovered optimal solution is not lost due to random genetic operations, an elite retention scale is set. In each generation iteration, the individual with the best fitness in the current generation is directly copied to the next generation, merged with the offspring, sorted by fitness, and the best individual is selected to form a new population. S534. Using a tournament selection strategy, individuals are randomly selected each time, and the one with the lowest fitness is selected to enter the next generation. This process is repeated to form a new candidate population. S54, Read the priority relationship set E Adjacency constraint matrix Decompose the weighted coefficients of the objective function Parameters, and simultaneously generate the initial disassembly sequence; S55. Extract the current generation best according to the elite retention strategy; generate the parent candidate set through tournament selection; complete the genetic operation of the genetic algorithm through selection, crossover mutation and constraint repair; evaluate the fitness of offspring; merge elites and offspring, select the best individuals to form the next generation elite; record the current generation best completion time, net profit and workstation utilization rate; use fast non-dominated sorting to screen out the Pareto optimal front to enter the next generation. S56. After reaching the maximum number of iterations, output the optimal disassembly sequence that minimizes disassembly time and maximizes disassembly profit.

7. A parallel dismantling sequence planning system for multiple scrap oil-immersed transformers, based on the batch scheduling-considered parallel dismantling sequence planning method for multiple scrap oil-immersed transformers as described in any one of claims 1-6, characterized in that, include: The module includes a data input module, a constraint decomposition and job time modeling module, a bufferless blocking flow batch scheduling modeling module, an optimization objective construction module, a multi-objective optimization solution module, and a visualization output module. The data input module is used to collect dismantling data of waste oil-immersed transformers, including transformer structure data, process parameter data and scheduling resource data. The dismantling constraint and operation time modeling module is used to establish an operation time model for the dismantling of waste oil-immersed transformers based on the collected dismantling data and in accordance with multi-dimensional composite constraints, and to calculate the effective operation time of each dismantling task. The bufferless blocking flow batch scheduling modeling module addresses the bufferless batch scheduling problem of multiple transformers at multiple serial workstations. Based on the batch occupancy constraints of the same workstation, the flow constraints of adjacent workstations, and the receiving constraints of downstream workstations, it establishes a batch decomposition time-spacing matrix for the entry, operation, departure, and blocking times of each transformer at each workstation. It also constructs a bufferless blocking flow recursive completion time model for multiple transformers at multiple workstations, calculates the entry and exit times of multiple devices in the pipeline, and calculates the scheduling table for each device in the pipeline. The optimization objective construction module is used to construct a multi-objective optimization model with the joint optimization objective of minimizing the total completion time of batch dismantling and maximizing the total profit of batch dismantling, and with workstation utilization and workstation load balancing as auxiliary optimization objectives. The multi-objective optimization solution module is used to solve the multi-objective optimization model based on the improved genetic algorithm, under the premise of satisfying all constraints, and output the decomposition sequence and workstation allocation scheme corresponding to the Pareto optimal front. The visualization output module is used to convert the scheduling time obtained from the optimization solution into a visualized batch dismantling Gantt chart, and output the dismantling task entry, operation, departure and blocking time of each workstation and each transformer, as well as the task content and the iterative convergence curve of the objective function.

Images