Ex-service power battery order matching and recombination scheduling cross-level cooperation method based on bilevel programming

Abstract

The invention relates to a double-layer planning-based decommissioned power battery order matching and recombination scheduling cross-level cooperation method, which comprises the following steps of: after finishing current matching decision iteration, issuing decided order matching parameters to a recombination scheduling planning subsystem; receiving a recombination scheduling decision parameter returned by the recombination scheduling planning subsystem in response to the corresponding order selection and distribution parameter, updating an order selection and distribution decision based on the recombination scheduling decision parameter and a preset NSGA-II algorithm, and generating an updated order selection and distribution parameter; according to a preset objective function, calculating the fitness corresponding to the updated order matching parameter, and when the fitness is not greater than a fitness threshold value, repeatedly performing order matching decision iteration based on the recombination scheduling decision parameter, the fitness and an NSGA-II algorithm returned by the recombination scheduling planning subsystem, so as to complete order matching decision iteration. And taking the received target recombination scheduling decision parameter and the received target order matching parameter as a cooperation result.

Description

Technical Field

[0001] This application relates to the technical field of cascade utilization of retired power batteries, in particular to a cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming. Background Art

[0002] Power batteries are widely used in fields such as electric vehicles. As the usage time goes by, their performance will gradually decay, and finally reach the retired state where they can no longer meet the vehicle's usage requirements. The retired power batteries still have a certain remaining capacity and can be used in scenarios such as energy storage. In related technologies, the ways of recycling and reusing power batteries include disassembly and recycling and cascade utilization. Cascade utilization is becoming increasingly popular in the recycling of power batteries because it can extend the life cycle value of the battery and achieve the full utilization of resources.

[0003] Different from traditional assembly manufacturing, in the scenario of cascade utilization of retired power batteries, there is a significant collaborative decision-making requirement in the link of component selection and recombination production of retired batteries. The parameters of multi-granularity remanufactured battery components after disassembly and recycling show high heterogeneity and discreteness (such as capacity, voltage, attenuation rate, internal resistance value, SOC state, etc.). Under the condition of meeting the voltage-current constraint conditions of the recombined product, the system will generate a selection combination plan in a multi-dimensional solution space. Since different selection plans will derive different process paths, recombination operation time, and equipment energy consumption costs, a process coupling relationship is formed between the selection decision of retired battery components and the recombination production scheduling. Moreover, the selection result directly affects the complexity of the recombination process and the equipment utilization rate. The efficiency of recombination production reflects the quality of the selection decision. The two constitute a multi-objective optimization problem with decision variable coupling, but they are at different decision levels. In related technologies, there is no implementation method that can coordinate the order selection decision at the planning level and the recombination scheduling at the execution level to find the optimal selection plan for retired power battery components, that is, there is no cross-level collaborative decision-making plan for battery recombination and selection.

[0004] Aiming at the problem that battery recombination and selection are not cross-level collaborative decision-making in related technologies, no effective solution has been proposed. Summary of the Invention

[0005] The embodiments of this application provide a cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming to at least solve the problem that battery recombination and selection are not cross-level collaborative decision-making in related technologies.

[0006] In a first aspect, an embodiment of the present application provides a cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming, including: after completing the current iteration of the order selection decision, sending the determined order selection parameters to the recombination scheduling planning subsystem; receiving the recombination scheduling decision parameters returned by the recombination scheduling planning subsystem in response to the corresponding order selection parameters, and based on the recombination scheduling decision parameters and a preset NSGA-II algorithm, updating the order selection decision to generate the updated order selection parameters, where the recombination scheduling decision parameters are generated by the recombination scheduling planning subsystem based on an improved adaptive catastrophe genetic algorithm and the corresponding order selection parameters; calculating the fitness corresponding to the updated order selection parameters according to a preset objective function, and determining whether the fitness is greater than a fitness threshold; in the case where it is determined that the fitness is not greater than the fitness threshold, repeatedly performing the order selection decision iteration based on the recombination scheduling decision parameters, the fitness, and the NSGA-II algorithm returned by the recombination scheduling planning subsystem until the target order selection parameters are generated, and using the received target recombination scheduling decision parameters and the target order selection parameters as the collaborative result, where the target recombination scheduling decision parameters are the recombination scheduling decision parameters used to generate the target order selection parameters.

[0007] Compared with the related art, the cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming provided by the embodiment of the present application adopts the following steps: after completing the current iteration of the order selection decision, sending the determined order selection parameters to the recombination scheduling planning subsystem; receiving the recombination scheduling decision parameters returned by the recombination scheduling planning subsystem in response to the corresponding order selection parameters, and based on the recombination scheduling decision parameters and a preset NSGA-II algorithm, updating the order selection decision to generate the updated order selection parameters; calculating the fitness corresponding to the updated order selection parameters according to a preset objective function, and determining whether the fitness is greater than a fitness threshold; in the case where it is determined that the fitness is not greater than the fitness threshold, repeatedly performing the order selection decision iteration based on the recombination scheduling decision parameters, the fitness, and the NSGA-II algorithm returned by the recombination scheduling planning subsystem until the target order selection parameters are generated, and using the received target recombination scheduling decision parameters and the target order selection parameters as the collaborative result. This method solves the problem in the related art that battery recombination and order selection are not cross-level collaboratively decided, and realizes the joint optimization of "order selection" and "recombination scheduling" that are closely related but at different decision levels, so as to achieve the cascade utilization of retired power batteries in a low-cost, high-efficiency, and sustainable manner.

[0008] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more concise and understandable. Description of the Drawings

[0009] The drawings described herein are used to provide a further understanding of the present application and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings: Figure 1 is a block diagram of the hardware structure of the terminal of the cross-level collaborative method for order selection, recombination, and scheduling of retired power batteries based on bilevel programming according to an embodiment of the present application; Figure 2 is a flowchart of the cross-level collaborative method for order selection, recombination, and scheduling of retired power batteries based on bilevel programming according to an embodiment of the present application; Figure 3 is a schematic diagram of the initial coding individual and the coding of alternative coding individuals according to an embodiment of the present application; Figure 4 is a schematic diagram of the crossover operation according to an embodiment of the present application; Figure 5 is a schematic diagram of the mutation operation for an embodiment of the present application; Figure 6 is a schematic diagram of the coding and decoding of the recombination scheduling code body for an embodiment of the present application; Figure 7 is the machProcArray base correspondence table for an embodiment of the present application; Figure 8 is a schematic diagram of a pair of parental chromosomes selected for the partially matched crossover operation according to an embodiment of the present application; Figure 9 is a schematic diagram of the invalid chromosome obtained by the exchange of a single matching crossover operation according to an embodiment of the present application; Figure 10 is the gene mapping table for an embodiment of the present application; Figure 11 is a schematic diagram of the offspring chromosome after gene matching according to an embodiment of the present application; Figure 12 is a schematic diagram of the insertion mutation for an embodiment of the present application. Detailed Embodiments

[0010] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be described and explained below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments provided in the present application without creative efforts belong to the scope of protection of the present application. In addition, it can also be understood that although the efforts made in such a development process may be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in the present application, some design, manufacturing or production changes based on the technical content disclosed in the present application are only conventional technical means and should not be understood as the content disclosed in the present application being insufficient.

[0011] In the present application, the mention of "embodiment" means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The appearance of this phrase at various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those of ordinary skill in the art explicitly and implicitly understand that the embodiments described in the present application can be combined with other embodiments without conflict.

[0012] Unless otherwise defined, the technical terms or scientific terms involved in the present application should be of the ordinary meaning understood by those with ordinary skills in the technical field to which the present application belongs. The words such as "a", "one", "a kind of", "the" and the like involved in the present application do not represent a quantity limitation and can represent a singular or plural number. The terms "including", "comprising", "having" and any variations thereof involved in the present application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device including a series of steps or modules (units) is not limited to the listed steps or units, but may further include unlisted steps or units, or may further include other steps or units inherent to these processes, methods, products or devices. The "multiple links" involved in the present application refer to two or more links. "And / or" describes the association relationship of associated objects and indicates that three relationships may exist. For example, "A and / or B" may represent: A exists alone, A and B exist simultaneously, and B exists alone. The terms "first", "second", "third", etc. involved in the present application are only used to distinguish similar objects and do not represent a specific order for the objects.

[0013] The relevant technologies used in the embodiments of the present application are described as follows: Aging Electric Vehicle Battery (AEVB) refers to lithium-ion, nickel-metal hydride and other types of batteries that have been phased out from electric vehicles or other electric devices. Their remaining capacity usually drops to 70%-80% of the initial capacity, making them unable to meet high-power demands, but still having secondary utilization value.

[0014] Cascade Utilization means that after retired power batteries (such as lithium-ion batteries phased out from electric vehicles) are tested, screened and reorganized, they are applied to other fields with lower performance requirements to extend their service life. When power batteries are retired from electric vehicles, their capacity usually still remains at 70%-80%. Although they cannot meet high-power demands, they can be adapted to scenarios such as energy storage and backup power supplies to achieve the maximum utilization of resources.

[0015] Order Matching means reasonably combining and matching the multi-granularity remanufactured battery components obtained from disassembly to meet the requirements of the capacity and other specifications of the cascade products, similar to the process of formulating a product BOM list.

[0016] Battery Reassembly means determining the reconfiguration process routes and reconfiguration times for different products according to the matching plan, and arranging the reconfiguration production sequence to meet the timely delivery of orders while improving production efficiency.

[0017] NSGA-II (Non-dominated Sorting Genetic Algorithm II) is a multi-objective optimization algorithm (MOEA) proposed by Kalyanmoy Deb et al. in 2002 for efficiently solving optimization problems with multiple conflicting objectives. It is an improved version of the classical NSGA. By introducing fast non-dominated sorting, crowding distance comparison operator and elitism strategy, it significantly improves the convergence and distribution of the algorithm.

[0018] Fast Non-dominated Sorting: First, the algorithm hierarchically sorts the individuals in the population according to the Pareto dominance relationship. If one solution is not inferior to another solution in all objectives and is better in at least one objective, it is said to dominate the latter. All non-dominated solutions form the first layer (Pareto front), and then the second layer of non-dominated solutions is continuously selected from the remaining solutions, and so on, forming multiple front levels (Front 1, Front 2, ...).

[0019] Crowding Distance Calculation: To maintain the diversity of the population, NSGA-II calculates the crowding distance of solutions within the same front layer to measure the distribution density of solutions in the objective space. Solutions with a high crowding distance are located in sparse regions, which helps to maintain the uniform distribution of the Pareto front and prevent the algorithm from converging prematurely to a local optimum.

[0020] Elitism Strategy: When generating offspring, NSGA-II combines the parent generation and the offspring, and then selects the optimal individuals to enter the next generation through non-dominated sorting and crowding distance comparison, ensuring that excellent solutions are not lost.

[0021] The following is a specific description of the embodiments of the present application: The method embodiments provided in this embodiment can be executed on a terminal, a computer, or a similar computing device. Taking running on a terminal as an example, Figure 1 is a hardware structure block diagram of the terminal of the cross-level collaborative method for retired power battery order selection and recombination scheduling based on bilevel programming in the embodiments of the present application. As Figure 1 shown, the terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, processing devices such as a microprocessor MCU or a field programmable gate array FPGA) and a memory 104 for storing data. Optionally, the above terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above terminal. For example, the terminal may further include more or fewer components than Figure 1 shown in the figure, or have a different configuration from Figure 1 shown in the figure.

[0022] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming in the embodiments of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above-mentioned method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely provided with respect to the processor 102, and these remote memories can be connected to the terminal through a network. Examples of the above-mentioned network include but are not limited to the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0023] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by a communication provider of the terminal. In one instance, the transmission device 106 includes a network adapter (abbreviated as NIC for Network Interface Controller), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 can be a radio frequency (abbreviated as RF) module, which is used to communicate with the Internet wirelessly.

[0024] This embodiment provides a cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming running on the above-mentioned terminal. Figure 2 It is a flowchart of the cross-level collaborative method for order selection and recombination scheduling of retired power batteries based on bilevel programming according to the embodiments of the present application, as Figure 2 shown, and the process includes the following steps:

[0025] Step S201, after completing the current selection decision iteration, send the selected order parameters to the recombination scheduling planning subsystem.

[0026] In this embodiment, an AEVB order selection and reorganization scheduling model is established using the Bilevel Interactive Optimisation (BIO) model framework. The execution entity for implementing the collaborative method of this application is the order selection (OM) decision-making system as the upper decision-making level, which focuses on formulating a selection plan for retired battery components for each order and forming the input of the reorganization scheduling planning subsystem to generate an optimal reorganization workshop scheduling plan with the shortest completion time, and achieving a balance among the performance of reorganized products, reorganization costs, and production efficiency. The reorganization scheduling planning subsystem, as the execution unit of the lower decision-making level and as the follower, makes a decision on the production order arrangement according to the corresponding selection plan after the execution entity generates the selection plan for retired battery components, that is, realizes the production goal under the constraint of limited resources, and optimizes according to the reorganization production completion time of the selection plan formulated by the upper decision-making level. Then, the execution unit of the lower decision-making level feeds back the decision-making workshop scheduling result to the execution entity for iterative optimization of the selection plan formulation. After that, the execution entity sends the corresponding order selection result (corresponding to the order selection parameters) to the execution unit of the lower decision-making level (corresponding to the reorganization scheduling planning subsystem) after iterative optimization. The execution unit of the lower decision-making level plans the workshop scheduling result according to the corresponding order selection parameters and judges the quality of the previously decided order selection plan through the reorganization scheduling decision parameters returned in response to the corresponding order selection parameters. Thus, multiple rounds of bilevel interactive optimization are repeated until the optimal order selection parameters and reorganization scheduling decision parameters are obtained. Therefore, after the execution entity completes the current selection decision iteration, it will send the decided order selection parameters to the reorganization scheduling planning subsystem of the lower decision-making level.

[0027] In this embodiment, the collaborative result includes the order selection and matching parameters determined by the upper decision-making level and the reorganized scheduling decision parameters generated by the lower decision-making level in response to the corresponding order selection and matching parameters. Among them, the order selection and matching parameters are based on order information (including product type, parameter specifications, reorganized product demand quantity, delivery date, and late delivery penalty), and under the constraints of meeting the order capacity of echelon products, energy requirements, etc., and meeting the multi-objective conditions of high consistency, long service life, low selection cost of the selected battery components, and high utilization rate of remanufacturing resources, a battery product reorganization matching plan is made, including the battery type, battery material type, battery component type (battery module, battery cell), and specifications (quantity) used in reorganizing and producing each order. It can be understood that in this embodiment, the selection of the battery component type for the order selection and matching parameters is based on considering minimizing the specification difference between the reorganized production product and the order demand product, minimizing the battery component selection cost (including reorganization cost, storage cost, and late delivery cost), and the storage cost of the battery components in the line-side warehouse of the reorganization workshop (overstock penalty cost, temporary storage cost); in this embodiment, the reorganized scheduling decision parameters depend on the corresponding order selection and matching parameters, and under the resource constraints of the corresponding production resources, including the inventory quantity, parameter specifications of various battery components, and the production line resources of the reorganization workshop (including the number of processes, the number of machines, and the line-side warehouse capacity), the production sequence arrangement plan for the reorganization of the battery products corresponding to the reorganization matching plan is realized to generate the optimal scheduling plan for each production line in each workshop to meet the timely delivery of the reorganized battery orders.

[0028] Step S202: Receive the reorganized scheduling decision parameters returned by the reorganized scheduling planning subsystem in response to the corresponding order selection and matching parameters, and based on the reorganized scheduling decision parameters and the preset NSGA-II algorithm, perform order selection and matching decision update to generate updated order selection and matching parameters. Among them, the reorganized scheduling decision parameters at least include the unit processing cost parameter of the process, the completion time parameter, and the order online time parameter, and are generated by the reorganized scheduling decision based on the improved adaptive mutation genetic algorithm and the process path parameter in the corresponding order selection and matching parameters.

[0029] In this embodiment, the OM decision-making system, which is the executor at the upper decision-making level, uses the NSGA-II algorithm and the reconfiguration scheduling decision parameters fed back by the reconfiguration scheduling planning subsystem to make decisions on the optimal battery component selection plan for all current orders; the reconfiguration scheduling planning subsystem, which is the executor at the lower decision-making level, uses the improved adaptive catastrophe genetic algorithm and the order selection parameters issued by the upper layer to make decisions on the optimal scheduling plan for each production line in each workshop; in this embodiment, before making decisions, the OM decision-making system and the reconfiguration scheduling planning subsystem will convert the corresponding decision variables and target variables into corresponding mathematical models to describe the order selection and reconfiguration scheduling processes of battery products. Among them, the OM decision-making system will use the objective function and constraint function of the corresponding upper-layer selection model to mathematize the order selection of reconfigured battery products, and the reconfiguration scheduling planning subsystem uses the objective function and constraint function of the corresponding lower-layer scheduling model to mathematize the production scheduling of reconfigured batteries in the workshop; in this embodiment, based on the collected order information, the OM decision-making system establishes a corresponding mathematical model to describe the selection process of the batteries to be reconfigured, that is, the process of formulating the selection plan of retired power battery components for each order. For example, for an order with a battery product demand, a battery component selection plan of selecting the same brand of battery modules + battery cells combination is adopted, with the goal of minimizing the specification difference between the reconfigured products and the products required by the order, minimizing the battery component selection cost (including reconfiguration cost, storage cost, late delivery cost), and the storage cost of battery components in the side warehouse of the reconfiguration workshop (excess penalty cost, temporary storage cost), and formulating the selection plan of retired power battery components for each order; in this embodiment, based on the collected production resources, the reconfiguration scheduling planning subsystem establishes a corresponding lower-layer mathematical model to describe the process of the optimal reconfiguration workshop scheduling with the shortest completion time (objective function), that is, the process of arranging the production order, so as to realize the production of the reconfigured battery products required for the corresponding order under the limited resource constraints.

[0030] In this embodiment, based on the collected order information, the OM decision-making system establishes a corresponding mathematical model to describe the selection process of the batteries to be reconfigured. Under the constraint conditions such as the capacity and energy demand of the cascade product orders, it is also necessary to meet multiple objectives such as high consistency, long service life, low selection cost, and high utilization rate of remanufacturing resources of the selected battery components, which belongs to the multi-objective polynomial difficult (NP-Hard) combinatorial optimization problem and is solved by the NSGA-II algorithm; in this embodiment, based on the collected production resources, the reconfiguration scheduling planning subsystem establishes a corresponding lower-layer mathematical model to describe the process of the optimal reconfiguration workshop scheduling with the shortest completion time, which belongs to the non-linear, multi-extremum, and multi-variable problem and is solved by the improved adaptive catastrophe genetic algorithm.

[0031] It should be noted that after the OM decision-making system at the upper decision-making level performs an iteration of order selection and matching decision-making and sends the corresponding order decision parameters to the reorganization scheduling and planning subsystem, the reorganization scheduling and planning subsystem will respond to the received order decision parameters. That is, based on the corresponding order selection and matching parameters (corresponding to the process path parameters therein) and the production resources available, the improved adaptive catastrophic genetic algorithm performs multiple decision-making iterations to generate reorganization scheduling decision parameters corresponding to the corresponding order selection and matching parameters, and obtains the optimal reorganization workshop scheduling plan with the shortest completion time. That is, in a cross-level collaborative interaction process, the OM decision-making system at the upper decision-making level performs an iteration of order selection and matching decision-making, and the reorganization scheduling and planning subsystem at the lower decision-making level will perform multiple corresponding decision-making iterations.

[0032] Step S203: Calculate the fitness corresponding to the updated order selection and matching parameters according to the preset objective function, and determine whether the fitness is greater than the fitness threshold. Among them, the objective function is constructed by coupling the AEVB product deviation parameter, the reorganization cost parameter, the late penalty cost parameter, and the line-side warehousing cost parameter. The reorganization cost parameter is associated with the unit processing cost parameter of the process, the late penalty cost parameter is associated with the completion time parameter, and the line-side warehousing cost parameter is associated with the order online time parameter.

[0033] In this embodiment, the objective function of the mathematical model corresponding to the upper-level selection and matching model couples the AEVB product deviation parameter (corresponding to the specification difference between the product produced by reorganization and the product required by the order), the reorganization cost parameter (including reorganization cost and storage cost), the late penalty cost parameter (including late delivery cost), and the line-side warehousing cost parameter (including the storage cost of battery components in the line-side warehouse of the reorganization workshop), and the corresponding parameter values will be used as the function values of the corresponding objective function and used to calculate the corresponding fitness; it can be understood that the function value of the objective function of the mathematical model corresponding to the lower-level scheduling model is also used to calculate the corresponding fitness; in this embodiment, after the OM decision-making system at the upper decision-making level completes an order selection and matching decision, it will calculate the fitness corresponding to the corresponding order selection and matching plan to determine the pros and cons of the order selection and matching plan corresponding to the decision, so as to guide whether the OM decision-making system performs decision iteration on the order selection and matching plan (corresponding to the order selection and matching parameters) until the corresponding order selection and matching plan meets the set requirements, that is, until the corresponding target order selection and matching parameters are generated.

[0034] Step S204, when it is determined that the fitness is not greater than the fitness threshold, repeat the order selection decision iteration based on the recombination scheduling decision parameters, fitness, and NSGA-II algorithm returned by the recombination scheduling planning subsystem until the target order selection parameters are generated, and use the received target recombination scheduling decision parameters and target order selection parameters as the collaboration result. The target recombination scheduling decision parameters are the rescheduling decision parameters used to generate the target order selection parameters.

[0035] In this embodiment, when it is determined that the fitness corresponding to the corresponding order selection parameters is not greater than the fitness threshold, it indicates that the order selection plan currently decided by the OM decision system is not the optimal battery component selection plan, and the decision on battery component selection needs to be continued. At this time, the OM decision system will make another order selection decision and send the corresponding order selection decision parameters to the recombination scheduling planning subsystem, so that the recombination scheduling planning subsystem makes multiple decisions on the optimal scheduling plan of each production line in the workshop, and generates the optimal production scheduling decision corresponding to the corresponding order selection parameters, that is, the corresponding recombination scheduling decision parameters.

[0036] In this embodiment, the OM decision system generates the optimal order selection parameters that meet the set requirements by performing multiple order selection decision iterations and collaborating across levels with the recombination scheduling planning subsystem at the lower decision level. It can be understood that meeting the set requirements includes, but is not limited to, the fitness of the optimal order selection parameters being greater than the fitness threshold, the number of iterations reaching the set number, and the difference in fitness between the order selection parameters of two consecutive times being within the preset change value range.

[0037] Through the above steps S201 to S204, after completing the current order matching decision iteration, the decision-making order matching parameters are sent to the reconfiguration scheduling planning subsystem; receive the reconfiguration scheduling decision parameters returned by the reconfiguration scheduling planning subsystem in response to the corresponding order matching parameters, and based on the reconfiguration scheduling decision parameters and the preset NSGA-II algorithm, perform order matching decision update to generate updated order matching parameters; according to the preset objective function, calculate the fitness corresponding to the updated order matching parameters, and determine whether the fitness is greater than the fitness threshold; in the case where it is determined that the fitness is not greater than the fitness threshold, repeat the order matching decision iteration based on the reconfiguration scheduling decision parameters, fitness, and NSGA-II algorithm returned by the reconfiguration scheduling planning subsystem until the target order matching parameters are generated, and use the received target reconfiguration scheduling decision parameters and target order matching parameters as the collaboration result to solve the problem in the related technology that battery reconfiguration and matching are not cross-hierarchy collaborative decisions, and realize the joint optimization of "order matching" and "reconfiguration scheduling" which are closely related but at different decision levels, so as to achieve the cascade utilization of retired power batteries in a low-cost, high-efficiency, and sustainable manner.

[0038] It should be noted that the objects "order matching" and "reconfiguration scheduling" faced by the embodiments of the present application are at different decision levels, and the "order matching" decision at the planning level significantly affects the "reconfiguration scheduling" decision at the scheduling level, and the "reconfiguration scheduling" decision result can be used to measure the quality of order matching. The two-layer optimization can be used to create an overall model of the inherent hierarchical structure between different levels of the supply chain, solve the interrelated decision-making network, and realize the integration of the planning and scheduling levels; the embodiments of the present application combine the principle of bilevel programming (BIO) to construct a two-stage cross-hierarchy collaborative decision-making of "order matching" and "reconfiguration scheduling"; in this embodiment, a nested two-layer optimization algorithm is used to solve the corresponding decision-making parameters, and the NSGA-II algorithm is used to solve the mathematical model corresponding to order matching to obtain the optimal battery component matching scheme for all current orders, and an improved adaptive catastrophic genetic algorithm is used to solve the mathematical model corresponding to reconfiguration scheduling to obtain the optimal scheduling scheme for each production line in each workshop to meet the timely delivery of customer orders.

[0039] In some of these embodiments, the reconfiguration scheduling decision parameters at least include the unit processing cost parameter of the process, the completion time parameter, and the order online time parameter, and the following steps are also implemented:

[0040] Step 21, determine all the reorganized orders corresponding to the order selection and matching parameters issued before the current upper-level decision, and encode and initialize the population for all the reorganized orders according to the upper-level optimization model corresponding to the preset two-layer optimization mathematical model to generate initial encoded individuals. The upper-level optimization model includes an AEVB product difference function, a reorganization cost function, a late penalty cost function, and a line-side warehousing cost function. The reorganization cost function is associated with the unit processing cost parameter of the process, the late penalty cost function is associated with the completion time parameter, and the line-side warehousing cost function is associated with the order online time parameter. The sub-encoding corresponding to the initial encoded individual is used to represent a reorganization product selection and matching decision.

[0041] In this embodiment, the NSGA-II algorithm is used to solve the order selection and matching problem. Therefore, it is necessary to perform genetic encoding on the relevant information corresponding to the decision of selecting and matching battery components, that is, perform chromosome gene encoding and population initialization according to the preset encoding form to form an initial population corresponding to the upper-level optimization model corresponding to the two-layer optimization mathematical model, that is, an initial population including multiple initial encoded individuals.

[0042] In this embodiment, one sub-encoding of the initial encoded individual corresponds to a reorganization product selection and matching decision, that is, the selection and matching of the battery components required for a reorganized order. For example, for an order, multiple battery modules are selected for reorganization, and the selected multiple battery modules are the corresponding selection and matching.

[0043] In this embodiment, when initializing the double-layer optimization mathematical model, the following settings are made for the upper-layer mathematical model to be used: 1. The demand of a customer order is for the same product; 2. The demand of a customer order is for the same product; 3. Each order can only be reorganized using one type of battery material, and batteries of the same brand and model should be used as much as possible; 4. Battery cells form a battery module in parallel, and battery modules form a battery pack in series; 5. Battery types are classified in sequence according to battery material type (such as ternary lithium, lithium iron phosphate, lithium titanate, etc.), battery component type (module or cell), and specifications (voltage and capacity); 6. To avoid battery damage due to violent transportation during outbound, etc., a certain proportion of redundancy is generally set according to the demand to ensure normal reorganization production; 7. The product categories produced by each workshop are different, and each workshop only produces one product; 8. The number of reorganization lines in each workshop is different; 9. The processes of all reorganization lines in the same workshop are the same; 10. A machine can only process one workpiece at a time; 11. Only one workpiece can be processed on one machine at the same time; 12. No workpiece has the right to be given priority in processing; 13. Once a workpiece starts processing, it cannot be interrupted; 14. The logistics time and logistics resource constraints between processes are not considered; 15. In the reorganization stage, all machines and workpieces are available at time zero, and all machines are in good working condition; 16. It is assumed that all battery components required for order reorganization are temporarily stored in the line-side warehouse of the reorganization workshop at the same moment; 17. It is assumed that the outbound time of battery components from the line-side warehouse is equal to the start production time of the order on the line, and the transfer time between the two is ignored.

[0044] In this embodiment, the mathematical model and the definitions of parameter symbols involved are as follows: i represents the order number, i ∈ {1, 2, 3, …, n}; G(i) represents the number of batteries required for reorganization of the i-th order; T i represents the expected delivery time of the battery product of the i-th order, represents the actual delivery time of the battery product of the i-th order; represents the unit delay penalty cost of the i-th order; s represents the battery material type, s ∈ {1, 2, ..., S}; a represents the battery parameter specification, a ∈ {1, 2, …, A}; q s,m,a represents the inventory quantity of the battery module of the a-th parameter specification in the s-th battery material type; q s,c,a represents the inventory quantity of the battery cell of the a-th parameter specification in the s-th battery material type; ah s,m,a represents the capacitance value of the battery module of the a-th parameter specification in the s-th battery material type; V s,m,a represents the voltage value of the battery module of the a-th parameter specification in the s-th battery material type; ah s,c,aRepresents the capacitance value of a battery cell with the a-th parameter specification in the s-th type of battery material; V s,c,a Represents the voltage value of a battery cell with the a-th parameter specification in the s-th type of battery material; Represents the parallel connection quantity when the battery cells selected in the i-th order are recombined into a battery module; Ah i Represents the capacitance of the AEVB required for the i-th order; Represents the capacitance of the AEVB actually produced for the i-th order; V i Represents the voltage of the AEVB required for the i-th order; Represents the voltage of the AEVB actually produced for the i-th order; α represents the redundancy ratio; RT represents the arrival time when the battery components required for all order recombinations are delivered to the recombination workshop; ST i Represents the start processing time of the i-th order on the recombination line; C ot Represents the unit temporary storage cost of the in-line warehouse in the recombination workshop; D r Represents the capacity upper limit of the in-line warehouse in the recombination workshop at time r; C oq Represents the unit penalty cost for overstock in the in-line warehouse of the final assembly workshop; j represents the recombination process label, j ∈ {1, 2,..., J}; t i,j Represents the processing time of the AEVB corresponding to the i-th order in the j-th recombination process, ; C j Represents the production cost per unit time for processing in the j-th recombination process; M j Represents the number of machines in the j-th recombination process; m represents the machine number in the j-th recombination process, m ∈ {1, 2,..., M j}; T i,j Represents the completion time of the i-th order in the j-th recombination process; E j Represents the earliest machine idle time in the j-th recombination process; Represents the out-of-warehouse time of the battery components required for the i-th order at the in-line warehouse; Represents the unit processing time of the battery module in the j-th recombination process; Represents the unit processing time of the battery cell in the j-th recombination process, Is a variable of 0 or 1, indicating whether the battery module is processed in the j-th recombination process. If so, = 1, otherwise = 0; Is a variable of 0 or 1, indicating whether the battery cell is processed in the j-th recombination process. If so, = 1, otherwise = 0; Represents the number of battery modules of the s-th type of battery material required for the AEVB corresponding to the i-th order; Represents the number of battery cells of the s-th type of battery material required for the AEVB corresponding to the i-th order; A variable that is 0 or 1, indicating whether the AEVB corresponding to the i-th order uses the battery module with the a-th parameter specification in the s-th type of battery material for reconfiguration production. If so, = 1, otherwise, = 0; A variable that is 0 or 1, indicating whether the AEVB corresponding to the i-th order uses the battery cell with the a-th parameter specification in the s-th type of battery material for reconfiguration production. If so, = 1, otherwise, = 0; A variable that is 0 or 1, indicating whether the AEVB corresponding to the i-th order uses battery components similar to the s-th type of battery material for reconfiguration production. If so, = 1, otherwise, = 0; A variable that is 0 or 1, indicating whether the i-th order on the m-th machine in the j-th reconfiguration process is the immediate predecessor operation of the -th order. If so, = 1, otherwise = 0; A variable that is 0 or 1, indicating whether the -th order is the first to be produced on the m-th machine in the j-th process. If so, = 1, otherwise, = 0; A variable that is 0 or 1, indicating whether the -th order is the last to be produced on the m-th machine in the j-th process. If so, = 1, otherwise, = 0; A variable that is 0 or 1, indicating whether the AEVB corresponding to the i-th order has been processed through the j-th process. If so, = 1, otherwise, = 0.

[0045] In this embodiment, for the OM decision-making system at the upper decision-making level, the objective function used to calculate the fitness value of a sub-code is: minimizing the specification difference between the reconfigured production product and the order-demand product and minimizing the battery component selection and matching cost , where C is the selection and matching cost, C p represents the reconfiguration cost, ; represents the late delivery cost, ; C s represents the storage cost of the battery components in the in-line warehouse of the reconfiguration workshop, C s= C1 + C2, where C1 represents the over - penalty cost, and C2 represents the temporary storage cost. ; The objective function set by the OM decision - making system is subject to the following constraint functions: 1. The outbound quantity of the battery module cannot exceed the current inventory, and the constraint function is: ; 2. The outbound quantity of the battery cell cannot exceed the current inventory, and the constraint function is: ; 3. The combined capacity of the battery modules and battery individuals selected for the AEVB of each order must be greater than or equal to the specified capacity required by the AEVB, and the constraint function is: ; 4. The combined voltage of the battery modules and battery individuals selected for the AEVB of each order must be greater than or equal to the specified voltage required by the AEVB, and the constraint function is: .

[0046] In some alternative implementation manners, according to the upper - layer optimization model corresponding to the preset two - layer optimization mathematical model, all re - organized orders are encoded and the population is initialized, including the following steps:

[0047] Step 1: Obtain the objective function and constraint parameters corresponding to the upper - layer optimization model, and obtain the production order information corresponding to each re - organized order, where the production order information includes the target information of the re - organized power battery.

[0048] Step 2: Select at least one battery combination matching information for each target information of the re - organized power battery from the preset battery combination matching information, where one battery combination matching information is used to characterize the composition of the battery materials and battery components corresponding to a re - organized power battery.

[0049] Step 3: Use the objective function and constraint parameters to solve and optimize the battery combination matching information corresponding to each target information of the re - organized power battery, generate the intended battery combination matching information corresponding to each re - organized power battery information, and perform two - dimensional integer encoding on the intended battery combination matching information in a preset encoding form to obtain the sub - encoding corresponding to the target information of the re - organized power battery.

[0050] Step 4: After determining the fitness value corresponding to each sub - encoding, integrate the sub - encodings corresponding to all re - organized orders into an initial encoding individual, and perform population initialization based on one initial encoding individual of the encoding to generate multiple initial encoding individuals.

[0051] In some preferred embodiments, referring to Figure 3 , the encoding process of the encoding individuals in the embodiments of the present application is described as follows:

[0052] According to the problem characteristics, each initial encoding individual is represented by a two - dimensional chromosome and all use the integer - encoding method. An initial encoding individual (referring toFigure 3 The vertical direction of the chromosome corresponding to P1 and P2 in it represents the order number Order, and the horizontal direction represents the number of battery types to be selected Type. The battery types are sequentially divided according to the battery material type (e.g., ternary lithium, lithium iron phosphate, lithium titanate), battery component type (e.g., battery module, battery cell), and specifications (e.g., voltage, capacity). In this embodiment, a certain row of the chromosome corresponds to a sub-code of an initial coding individual; Refer to Figure 3 , the ternary lithium battery material types include M1, C11, and C12, the lithium iron phosphate battery material type includes C2, and the lithium titanate battery materials include M3 and C3. Considering Figure 3 the P1 coding individual (corresponding to a chromosome) described in, M1, C11, and C12 belong to the same material type. M1 and C11, C12 belong to different battery component types under the same material type. M1 is a battery module, and C11, C12 are battery cells with different specifications. Correspondingly, C2 belongs to the battery cell of the second material type, and M3, C3 belong to different battery component types of the third material type; It should be noted that for an order, only battery components of the same material type can be used. For example, in Figure 3 the coding individual of P2 described in, for order 1 (#1), when there is data in the first three columns, the 4th - 6th columns will all be processed as 0. Similarly, for order 2, when there is data in the 5th - 6th columns, the first 4 columns will all be processed as 0. Such a coding method encodes the chromosomes of the population and initializes the population, which can easily meet the constraint of only using battery components of the same material type and can be more flexibly and effectively used for subsequent processing.

[0053] Step 22, based on the NSGA-II algorithm, the upper-layer optimization model, and the fitness corresponding to all initial coding individuals, perform non-dominated sorting of the NSGA-II algorithm on multiple initial coding individuals, and perform genetic evolution operations on the candidate coding individuals that have completed non-dominated sorting to generate the current coding individuals corresponding to the current upper-layer decision. Among them, the fitness is determined according to the sub-individual objective values calculated by the objective functions of the coupling AEVB product difference function, recombination cost function, deferred penalty cost function, and line-side warehousing cost function for the sub-codes corresponding to the initial coding individuals. The genetic evolution operations include tournament selection operation, multi-point crossover operation, and non-uniform mutation operation.

[0054] In this embodiment, during the process of solving the order selection and matching decision using the NSGA-II algorithm, the function value of the objective function that couples the AEVB product deviation parameters (corresponding to the specification differences between the products produced by recombination and the products required by the order), the recombination cost parameters (including recombination cost and storage cost), the late delivery penalty cost parameters (including late delivery cost), and the in-line warehouse cost parameters (including the storage cost of battery components in the in-line warehouse of the recombination workshop) is used as the fitness value of a sub-code, and the fitness values of multiple sub-codes are aggregated into the fitness corresponding to an initial code individual; in this embodiment, the non-dominated sorting of the initial code individual and the genetic evolution operation on the individuals after the non-dominated sorting are an update of the order selection and matching decision. The update of the order selection and matching decision includes the number and specifications (voltage and capacitance) of the battery modules and battery cells of a certain type of battery material (e.g., ternary lithium, lithium iron phosphate) used for the battery components selected for each order. The decision-making objectives include the specification differences between the AEVB produced by recombination and the AEVB required by the order, minimizing the selection cost, recombination cost, late delivery cost, and storage cost of battery components in the in-line warehouse of the recombination workshop, including overage penalty cost and temporary storage cost.

[0055] It can be understood that the non-dominated sorting performed by the NSGA-II algorithm in the embodiments of the present application is a well-known technology to those skilled in the art and does not constitute a limitation on the clarity of the technical solution. That is, based on the existing NSGA-II algorithm, non-dominated sorting can be performed on the initial code individual and its sub-codes, and then the corresponding candidate code individuals can be obtained, and then the relevant genetic evolution operations adopted in this embodiment can be performed. Specifically, it is to perform tournament selection operation, multi-point crossover operation, and non-uniform mutation operation.

[0056] Step 23: Use the order selection and matching parameters generated by decoding the current code individual as the updated order selection and matching parameters.

[0057] Through the above steps 21 to 23, the update of the order selection and matching decision at the genetic upper decision level is realized.

[0058] In some of these embodiments, the genetic evolution operation on the candidate code individuals after the non-dominated sorting is realized through the following steps:

[0059] Step 31: Randomly select any two candidate code individuals from the multiple alternative code individuals obtained by performing the tournament selection operation on multiple candidate code individuals as the intended code body group. After selecting multiple sub-codes to be crossed from all the sub-codes corresponding to the two alternative code individuals in the intended code body group, cross the sub-codes selected from the corresponding two alternative code bodies to generate two corresponding first-generation sub-code bodies.

[0060] In this embodiment, according to the fitness values of candidate coding individuals, a tournament selection operation is performed to select some individuals to be crossed (corresponding to a preset number of alternative coding individuals) from the parental population (corresponding to all candidate coding individuals in the current iteration). Individuals with higher fitness values have a greater probability of being selected. In this embodiment, after selecting the alternative coding individuals, a multi-point crossover operation is performed using the order-based multi-point crossover method (Multi-Point Crossover, abbreviated as MPX). Through the parents P1 and P2 (refer to Figure 3 ), the offspring C1 and C2 are generated by crossover (refer to Figure 4 ). The multi-point crossover process is as follows: (1) First, randomly select multiple points in the vertical coding of the chromosome and find the order numbers corresponding to these points; then, according to the obtained order numbers, select the corresponding battery type combinations assigned to these orders in the horizontal coding of P1 and P2 chromosomes respectively; 3. Finally, exchange the coding of the corresponding battery type combinations of the selected orders in P1 and P2 to generate the crossed offspring C1 and C2.

[0061] Step 32: Select the sub-codings to be mutated from all the sub-codings of the two corresponding first offspring coding bodies in each group of intended offspring coding body groups, and determine at least one intended coding in the selected sub-codings, where the intended coding is used to represent the coding with a non-zero corresponding coding value.

[0062] Step 33: Randomly increase or decrease the at least one intended coding according to a preset mutation intensity control number, and use the mutated coding as the new coding of the corresponding sub-coding to generate a second offspring coding body corresponding to each first sub-coding body, obtaining the coding individuals corresponding to the current upper-level decision.

[0063] In this embodiment, non-uniform mutation (corresponding to the mutation of non-zero positions) is used to mutate all the sub-codings of the first offspring coding body. Through the parents C1 and C2 (refer to Figure 4 ), the offspring D1 and D2 are generated by mutation (refer to Figure 5 ). In this embodiment, for the non-zero positions in a chromosome, mutation processing is performed. First, find the non-zero positions in the chromosome, and randomly increase or decrease a group of integers controlled by the mutation intensity for the non-zero positions, thereby achieving the mutation effect.

[0064] Through the above steps 31 to 33, a genetic evolution operation is implemented on the candidate coding individuals that have completed non-dominated sorting.

[0065] In some of these embodiments, before generating the target order matching parameters, the following steps are further implemented:

[0066] Step 41: Determine whether the order selection and matching decision iteration meets the preset iteration termination condition after the current upper-level decision. The iteration termination condition includes at least one of the following: the fitness difference between the current upper-level decision iteration and the fitness of the current coding individual corresponding to the previous upper-level decision iteration is less than the population fitness change threshold, or the number of times of the order selection and matching decision iteration exceeds the preset iteration number threshold.

[0067] Step 42: When it is determined that the iteration termination condition is met, use the order selection and matching parameters generated by the current upper-level decision as the target order selection and matching parameters.

[0068] Step 43: When it is determined that the iteration termination condition is not met, perform the next order selection and matching decision iteration based on the recombination scheduling decision parameters, fitness, and NSGA-II algorithm returned by the recombination scheduling planning subsystem.

[0069] Through the above steps 41 to 43, it is determined whether the target order selection and matching parameters are obtained.

[0070] In some of these embodiments, for the recombination scheduling planning subsystem, the following steps are adopted to perform the operation of recombination scheduling decision based on the improved adaptive catastrophic genetic algorithm and the corresponding order selection and matching parameters:

[0071] Step 51: After receiving the corresponding order selection and matching parameters, obtain all the recombination orders from the corresponding order selection and matching parameters, and determine the target operation process parameters specified for each recombination order.

[0072] In this embodiment, the recombination scheduling workshop sets a target number of target operation processes. However, due to different battery component selections in the order selection and matching parameters obtained from order selection and matching, there may be a situation where the order of one or more target operation processes is adjusted (not performed). For example: one order selection and matching parameter 1 corresponds to the need for target operation processes 1, 2, 3, 4, 5, 6, 7, and one order selection and matching parameter 2 corresponds to the need for target operation processes 1, 3, 6, 7. For the target operation processes required by a certain order selection and matching parameter, they are determined when the order selection and matching parameter is received. However, for which order selection and matching parameter's corresponding order is arranged first for a certain target operation process can be scheduled and decided. For example: the AEVB required to be produced by order 1 associated with order selection and matching parameter 1 needs to operate on target operation process 3, and the AEVB required to be produced by order 2 associated with order selection and matching parameter 2 also needs to operate on target operation process 3. Then, for this target operation process 3, the production of order 2 can be arranged first, and then the production of order 1, or the production of order 1 can be arranged first, and then the production of order 2. Therefore, the improved adaptive catastrophic genetic algorithm and the corresponding order selection and matching parameters can be used for recombination scheduling decision.

[0073] In this embodiment, after obtaining all the reorganized orders and determining the target operation process parameters specified for each reorganized order, decision-making planning can be performed on the target operation process corresponding to each reorganized order to produce the corresponding reorganized scheduling, that is, the arrangement of the production order of the reorganized order at each operation process station.

[0074] Step 52: Based on the lower-layer optimization model corresponding to the two-layer optimization mathematical model, perform integer coding and initialization on the reorganized orders and the target operation process parameters according to the preset coding rules to generate multiple reorganized scheduling coding bodies. Among them, the lower-layer optimization model includes a makespan objective function and multiple target sub-constraint parameters. The sub-coding corresponding to the reorganized scheduling coding body is generated by coding according to the makespan objective function and the target sub-constraint parameters, and is used to represent the sorting of the reorganized orders associated with a target operation process parameter.

[0075] In this embodiment, the reorganized scheduling planning subsystem uses an improved adaptive catastrophe genetic algorithm (ImproveAdaptive Catastrophe Genetic Algorithm, abbreviated as IACGA) to solve the reorganized scheduling decision. Furthermore, genetic coding is required for the arrangement of the operation processes for the reorganized operation of the battery components, that is, chromosome gene coding and population initialization are performed according to the preset coding form to form the initial population corresponding to the lower-layer optimization model corresponding to the two-layer optimization mathematical model, that is, multiple reorganized scheduling coding bodies.

[0076] In this embodiment, after the reorganized scheduling coding body is coded, the reorganized scheduling planning subsystem calculates the objective value of the makespan objective function corresponding to the sub-coding body of the reorganized scheduling coding body, and calculates the scheduling fitness of the corresponding reorganized scheduling coding body. In this embodiment, the makespan objective function is: , and the constraint functions corresponding to the multiple target sub-constraint parameters are as follows: 1. The constraint that any order can only have one immediate predecessor order on all machines in the same process, and the constraint function is: ; 2. The constraint that any order can only be the immediate predecessor order of one job on all machines in the same process, and the constraint function is: ; 3. There is no immediate predecessor and successor relationship between the orders at the same position on the same machine, and the functional formula is: ; 4. The constraint between the completion times of the sequential operations of the current machine, and the constraint function is ; 5. The start processing time of the i-th order, ; 6. The start processing time of the i-th order is equal to the material out time, and the constraint function is: ; 7. The constraint on the completion time between the front and back processes of a specific operation, and the functional formula is ; 8. The start time of the \(i\)th order in the current process is equal to the maximum value of the completion time of the previous process and the earliest machine idle time in the \(j\)th process. The constraint function is: ; 9. The delivery date constraint of the order, the function is: .

[0077] In some preferred embodiments, referring to Figure 6 and Figure 7 , the encoding and decoding processes of the re - scheduling coding body of the embodiments of the present application are described as follows: An integer coding method based on operations is adopted. Each individual is represented by a one - dimensional chromosome. The coding length is only related to the number of jobs \(n\) and the number of processes \(m\). The coding length is \(n\times m\). The coding content is integer values between 1, 2, …, \(n\times m\). For example, considering Figure 6 the L1 individual depicted, which consists of 4 jobs and 3 processes. Thus, the coding length is 12, and the coding content is values between 1, 2, …, 12; referring to Figure 6 , after decoding L1, the corresponding shop floor scheduling plan will be obtained (refer to the Job and Process parts in Figure 6 ).

[0078] In this embodiment, referring to Figure 7 , in order to obtain the scheduling arrangement of the job shop at each moment, it is necessary to convert the chromosome coding into a specific shop floor scheduling plan through decoding. The final scheduling plan array is , where \(L = n\times m\). The specific decoding steps are:

[0079] Step 1. Obtain the current chromosome nowChrome, with a length of \(L=n\times m\).

[0080] Step 2. Define the detailed scheduling plan array schedule[L][5]. The scheduling plan consists of five columns of data: jobId, machId, processId, startTime, and endTime.

[0081] Step 3: Let the current process \(j = 1\), and \(j\in(1, 2,\ldots, J)\) for traversal.

[0082] Step 4. According to the "basic correspondence table", find the codes of all jobs of process \(j\) from nowChrome, and retain their order in nowChrome. Assume it is procSeq, and find the corresponding job code jobSeq. The corresponding calculation logic is: first find the numbers in nowChrome that are less than or equal to \(m\times i\), and then find the numbers greater than \(m\times(i - 1)\) from the above sequence, denoted as procSeq. Take the remainder of (the number - 1) divided by \(m\) and add 1 to get jobSeq.

[0083] Step 5. Arrange the operations on the machines according to the operation coding sequence jobSeq of the current process j (priority rule).

[0084] In this embodiment, it is necessary to obtain the numbers nowMachIds of the currently available machines according to the process, and use a similar "basic correspondence table". Assuming that there are 2 machines for process 1, 3 machines for process 2, and 2 machines for process 3, the array machProcArray can be expressed as Figure 7 。

[0085] Step 53. Based on the preset catastrophic adaptive genetic algorithm, the lower-layer optimization model, and the scheduling fitness corresponding to the recombination scheduling code bodies participating in each recombination scheduling iteration, perform iterative genetic evolution operations and population catastrophe operations on multiple corresponding recombination scheduling code bodies until multiple intended recombination scheduling code bodies are generated. The scheduling fitness is calculated and determined according to the completion time objective function corresponding to the sub-codes corresponding to the recombination scheduling code bodies.

[0086] In this embodiment, the process of performing population iteration on a recombination scheduling code body and its corresponding sub-codes is a decision update for the production sequence arrangement of the recombination shop scheduling; it can be understood that the dynamic selection and adaptive crossover and mutation performed by the catastrophic adaptive genetic algorithm are all well-known techniques to those skilled in the art and do not constitute a limitation on the clarity of this technical solution. At the same time, it is also feasible to trigger a catastrophe according to the preset catastrophe rule to guide the corresponding genetic evolution operation iteration of the adaptive genetic algorithm; in this embodiment, according to the fitness of the population individuals in each iteration, it is determined whether to perform a catastrophe and whether the population individuals after iteration (corresponding to candidate code bodies) meet the set requirements, that is, whether they are intended recombination scheduling code bodies. Through multiple catastrophe operations and genetic evolution operations, the fitness of the individuals in the corresponding population meets the set requirements, and then multiple intended recombination scheduling code bodies are obtained.

[0087] Step 54. Obtain the target recombination scheduling code body from multiple intended recombination scheduling code bodies, and after decoding the target recombination scheduling code body, generate the recombination scheduling decision parameters corresponding to the current lower-layer decision.

[0088] In some alternative embodiments, obtaining the target recombination scheduling code body from multiple intended recombination scheduling code bodies is achieved through the following steps:

[0089] Step 54-1. Determine the fitness corresponding to each intended recombination scheduling code body.

[0090] Step 54-2: Select the intention recombination scheduling code body with the highest fitness from multiple intention recombination scheduling code bodies to obtain a target recombination scheduling code body, and extract all sub-codes of the target recombination scheduling code body to obtain target sub-codes.

[0091] Step 54-: Decode the recombination scheduling decision information corresponding to all target sub-codes, and use the recombination decision information corresponding to all target sub-codes as the recombination scheduling decision parameters returned in response to the corresponding order selection parameters.

[0092] Through the above steps 51 to 54, the recombination scheduling planning subsystem realizes the operation of making recombination scheduling decisions based on the improved adaptive catastrophic genetic algorithm and the corresponding order selection parameters.

[0093] In some embodiments, for the recombination scheduling planning subsystem, the following steps are adopted to perform iterative operations of genetic evolution operations and population catastrophe operations on multiple corresponding recombination scheduling code bodies:

[0094] Step 61: According to the fitness of multiple recombination scheduling code bodies participating in the current recombination scheduling iteration, use the elitist retention strategy to select a preset number of alternative scheduling code bodies from multiple recombination scheduling code bodies participating in the current recombination scheduling iteration.

[0095] In this embodiment, according to the fitness value of the recombination scheduling code body, the elitist retention strategy is adopted to retain the top 5% of excellent individuals to implement the selection operation.

[0096] Step 62: Based on the corresponding adaptive crossover probability and adaptive mutation probability, perform partially matched crossover operations and insertion mutation operations on all alternative scheduling code bodies in sequence to generate multiple first coding individuals. Among them, the recombination scheduling code body corresponding to the current genetic evolution operation iteration includes the first coding individuals. The adaptive crossover probability is determined based on the crossover probability function corresponding to the catastrophic adaptive genetic algorithm, and the adaptive mutation probability is determined based on the mutation probability function corresponding to the catastrophic adaptive genetic algorithm.

[0097] In this embodiment, for the selected preset number of alternative scheduling code bodies, crossover operations are first performed based on an adaptively adjusted crossover probability to generate new individuals. In this embodiment, the crossover probability is adaptively and dynamically adjusted according to the fitness value of the population and the evolutionary stage. In the initial stage of evolution, a higher crossover probability helps to maintain the diversity of the population, that is, a higher crossover rate is used for individuals with poorer fitness to increase the exploration range. In the later stage of evolution, a lower crossover probability is beneficial for local optimization, that is, a lower crossover rate is used for individuals with higher fitness to protect excellent genes. In this embodiment, after generating new individuals through crossover, mutation operations are performed on the genes of the new individuals based on an adaptively adjusted mutation probability to generate multiple first coding individuals. In this embodiment, the mutation probability can also be adaptively adjusted according to the evolutionary characteristics of the population and the fitness value of the individual. In the initial stage of evolution, a higher mutation probability helps to explore the entire search space, that is, a higher mutation rate is used for individuals with poorer fitness to increase the exploration range. In the later stage of evolution, a lower mutation probability helps to focus on local search to find the optimal solution, that is, a lower mutation rate is used for individuals with higher fitness to protect excellent genes. It can be understood that as the evolution progresses, the algorithm gradually approaches the optimal solution. At this time, the crossover probability and mutation probability should be appropriately reduced to protect the obtained excellent genotypes from being destroyed, and at the same time, allow fine-grained search to improve the quality of the solution. Therefore, in this embodiment, the crossover probability function and mutation probability function are set as monotonically decreasing functions. For a minimization problem, the smaller the individual fitness value, the smaller the current crossover probability and mutation probability. In this embodiment, the set crossover probability function and mutation probability function are used to adaptively adjust the crossover probability and mutation probability.

[0098] In this embodiment, with reference to Figures 8 to 12 , the process of sequentially performing partial matching crossover operations and insertion mutation operations on all alternative scheduling code bodies in the embodiments of the present application is described as follows:

[0099] The partial matching crossover ensures that each gene in a chromosome appears only once. Through this crossover strategy, no duplicate genes will appear in a chromosome. By randomly selecting two crossover points to determine the crossover region, generally two invalid chromosomes will be obtained after crossover, and individual genes may appear repeatedly. To repair the chromosomes, the matching relationship of each chromosome can be established within the crossover region, and then this matching relationship can be applied to the duplicate genes outside the crossover region to eliminate conflicts.

[0100] For the partial matching crossover operation, the following steps are included:

[0101] Step 1: Randomly select a pair of chromosomes (parent generation, reference Figure 8The start and end positions of several genes in P1 and P2) (the selected positions on the two chromosomes are the same). Assume that the 4th position is selected (corresponding to Figure 8 "2" of P1 and "7" of P2 pointed by the arrow in Figure 8 and the 8th position (corresponding to

[0102] "1" of P1 and "8" of P2 pointed by the arrow in Figure 9 P1 and P2).

[0103] Step 2: Exchange the positions of these two groups of genes to obtain two invalid chromosomes, and individual genes appear repeatedly (refer to Figure 10 P1 and P2 in Figure 11 ).

[0104] For the insertion mutation operation, new chromosomes are generated by inserting the genes at a certain locus into other loci of the chromosome, including the following steps:

[0105] Step 1: Randomly select a chromosome 𝑃1 (refer to Figure 12 ).

[0106] Step 2: Randomly select a gene on chromosome 𝑃1 and randomly move this gene to other loci in the chromosome to generate the offspring chromosome C1.

[0107] In this embodiment, first, generate two random numbers pos1 and pos2 between 1 and L (for example, Figure 12 "7" pointed by the arrow in

[0108] ). Judge the magnitudes of pos1 and pos2. If pos1 < pos2, move the genes between pos1 and pos2 forward by one position, and then assign pos1 to pos2; if pos1 > pos2, move the genes between pos1 and pos2 backward by one position, and then assign pos1 to pos2.

[0109] Step 63: Among all the recombinant scheduling code bodies generated in the previous genetic evolution iteration, select the recombinant scheduling code body with the optimal fitness to obtain the first optimal individual corresponding to the previous genetic evolution iteration. Then, from multiple first coding individuals, select the first coding individual with the optimal fitness to obtain the second optimal individual corresponding to the current genetic evolution iteration, and determine whether the fitness of the second optimal individual is higher than that of the first optimal individual.

[0110] In this embodiment, the optimal individual in the initial population (corresponding to all the recombinant scheduling code bodies generated in the previous genetic evolution iteration), which is the first optimal individual, is saved as the first local optimal individual. After several genetic operations on the population, whether the fitness of the second optimal individual is higher than that of the first optimal individual is used as the judgment condition for determining whether a better individual is generated and whether to perform a catastrophe operation. If the catastrophe determination condition is met, the catastrophe operation is performed.

[0111] Step 64: According to the judgment result, select the target optimal individual corresponding to the current genetic evolution iteration from the second optimal individual and the first optimal individual, and perform a catastrophe operation on multiple first coding individuals in a preset catastrophe manner to generate multiple second coding individuals. Among them, the recombinant scheduling code bodies corresponding to the current population catastrophe operation iteration include the second coding individuals.

[0112] In some alternative embodiments, for the recombinant scheduling planning subsystem, the following steps are adopted to perform the operation of performing a catastrophe operation on multiple first coding individuals in a preset catastrophe manner to generate multiple second coding individuals.

[0113] Step 641: When it is determined that the fitness of the second optimal individual is higher than that of the first optimal individual, select a preset number of first coding individuals in descending order of fitness to obtain elite coding individuals, and initialize all elite coding individuals based on the lower-layer optimization model to generate corresponding multiple second coding individuals.

[0114] In this embodiment, if the optimal individual in the current population (corresponding to the second optimal individual) is better than the previously updated local optimal individual (corresponding to the first optimal individual), then update the current local optimal individual and perform a type of catastrophe operation, that is, select a preset number of first coding individuals in descending order of fitness to obtain elite coding individuals, and initialize all elite coding individuals to generate corresponding multiple second coding individuals. In this embodiment, it is preferably to initialize each individual in the elite class of the current population (corresponding to the elite coding individuals) one by one.

[0115] Step 642, when it is determined that the fitness of the second optimal individual is not higher than that of the first optimal individual, select the first encoded individuals with a set cataclysm scale number in ascending order of fitness to obtain the eliminated encoded individuals. After removing all the eliminated encoded individuals from the multiple first encoded individuals, add multiple randomly generated new encoded individuals to the remaining first encoded individuals to obtain the corresponding multiple second encoded individuals, where the number of new encoded individuals is equal to the set cataclysm scale number.

[0116] In this embodiment, if the optimal individual in the current population (corresponding to the second optimal individual) is inferior to the locally optimal individual saved in the previous cataclysm (corresponding to the first optimal individual), it is considered that the search direction of the current population has deviated, and a type-II cataclysm operation is performed, that is, eliminate M individuals with poor fitness in the population and randomly generate M new individuals to join the current population to improve the diversity of the population; the current cataclysm scale M is not constant but decreases with the increase of the number of iterations to ensure the stability of the algorithm in the later stage. In this embodiment, the following formula is used to calculate the current cataclysm scale M: M is the current cataclysm scale, is the preset cataclysm scale, λ is the control parameter, is the current number of iterations, is the maximum number of iterations.

[0117] In some of these embodiments, for the recombinant scheduling planning subsystem, after obtaining the corresponding multiple second encoded individuals, the following steps are further performed:

[0118] Step 71, determine whether the current iteration satisfies the preset iteration termination condition after the current population cataclysm operation iteration, where the iteration termination condition includes at least one of the following: the fitness difference between the current population cataclysm operation iteration and the second encoded individuals corresponding to the previous population cataclysm operation iteration is less than the population fitness change threshold, and the number of the current population cataclysm operation iteration exceeds the preset iteration number threshold.

[0119] Step 72, when it is determined that the iteration termination condition is satisfied, use the multiple second encoded individuals as multiple intended recombinant scheduling code bodies.

[0120] Step 73, when it is determined that the iteration termination condition is not satisfied, perform iterative genetic evolution operations and population cataclysm operations on the multiple second encoded individuals until multiple intended recombinant scheduling code bodies are obtained.

[0121] In this embodiment, it is determined whether the algorithm terminates according to preset termination conditions (such as the number of iterations, population fitness change). If the termination conditions are met, the optimal solution is output (that is, multiple intended recombination scheduling coding bodies are output); if not, the next iteration is entered until the optimal solution is output.

[0122] In some of these embodiments, for the recombination scheduling planning subsystem, before generating multiple first coding individuals, the following steps are further implemented:

[0123] Step 81: According to the fitness of the alternative scheduling coding bodies, determine the average fitness and the minimum fitness corresponding to all alternative scheduling coding bodies, and determine whether the average fitness is less than the minimum fitness.

[0124] Step 82: In the case where it is determined that the average fitness is less than the minimum fitness, use the preset maximum crossover probability and maximum mutation probability as the corresponding adaptive crossover probability and adaptive mutation probability, respectively.

[0125] Step 83: In the case where it is determined that the average fitness is greater than the minimum fitness, use the corresponding crossover probability function and mutation probability function to decrease the maximum crossover probability and the maximum mutation probability respectively, and use the first crossover probability and the first mutation probability obtained after the decrease as the corresponding adaptive crossover probability and adaptive mutation probability, respectively.

[0126] In this embodiment, the following crossover probability function and mutation probability function are used to adaptively adjust the crossover probability and the mutation probability. Specifically, Crossover probability function Mutation probability function Among them, respectively represent the lower bound and the upper bound of the crossover probability, f min represents the minimum fitness value in the first coding individuals corresponding to the previous genetic evolution operation iteration, f avg represents the average fitness of the alternative scheduling coding bodies, represents the minimum fitness value among the alternative scheduling coding bodies, respectively represent the upper bound and the lower bound of the mutation probability, f m represents the fitness value of the mutant individual (the alternative coding individual that has completed the current crossover operation); based on the characteristics of the sigmoid function ( ), the crossover probability function and the mutation probability function are monotonically decreasing functions, and the crossover probability and the mutation probability are monotonically decreasing.

[0127] This embodiment also provides a cross - hierarchical collaborative system for AEVB order selection and reorganization scheduling based on bilevel programming, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any of the above - mentioned method embodiments.

[0128] Optionally, the above - mentioned collaborative system may further include a transmission device and an input - output device. Among them, the transmission device is connected to the above - mentioned processor, and the input - output device is connected to the above - mentioned processor.

[0129] Optionally, in this embodiment, the above - mentioned processor may be configured to execute the following steps through a computer program:

[0130] S1. After completing the current selection decision iteration, send the selected order parameters determined by the decision to the reorganization scheduling planning subsystem.

[0131] S2. Receive the reorganization scheduling decision parameters returned by the reorganization scheduling planning subsystem in response to the corresponding order selection parameters. Based on the reorganization scheduling decision parameters and the preset NSGA - II algorithm, update the order selection decision, and generate updated order selection parameters.

[0132] S3. According to the preset objective function, calculate the fitness corresponding to the updated order selection parameters, and determine whether the fitness is greater than the fitness threshold.

[0133] S4. In the case where it is determined that the fitness is not greater than the fitness threshold, repeat the order selection decision iteration based on the reorganization scheduling decision parameters, fitness, and NSGA - II algorithm returned by the reorganization scheduling planning subsystem until the target order selection parameters are generated. And use the received target reorganization scheduling decision parameters and target order selection parameters as the collaborative result. The target reorganization scheduling decision parameters are the reorganization scheduling decision parameters used to generate the target order selection parameters.

[0134] It should be noted that the specific examples in this embodiment can refer to the examples described in the above - mentioned embodiments and optional implementation manners, and will not be repeated here.

[0135] In addition, in combination with the above - mentioned cross - hierarchical collaborative method for retired power battery order selection and reorganization scheduling based on bilevel programming in the embodiment, the embodiment of the present application can provide a storage medium to implement. A computer program is stored on the storage medium; when the computer program is executed by a processor, it implements any one of the above - mentioned cross - hierarchical collaborative methods for retired power battery order selection and reorganization scheduling based on bilevel programming.

[0136] Those skilled in the art should understand that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as within the scope described in this specification.

[0137] The above embodiments only express several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A cross - level collaborative method for order selection, matching and restructuring scheduling of retired power batteries based on bilevel programming, characterized in that Including: After completing the current order matching decision iteration, send the order matching parameters determined by the decision to the reorganization scheduling planning subsystem; Receive the reorganization scheduling decision parameters returned by the reorganization scheduling planning subsystem in response to the corresponding order matching parameters, and based on the reorganization scheduling decision parameters and the preset NSGA-II algorithm, perform order matching decision update to generate the updated order matching parameters. Among them, the reorganization scheduling decision parameters are generated by the reorganization scheduling planning subsystem based on the improved adaptive catastrophe genetic algorithm and the corresponding order matching parameters; According to the preset objective function, calculate the fitness corresponding to the updated order matching parameters, and determine whether the fitness is greater than the fitness threshold; In the case where it is determined that the fitness is not greater than the fitness threshold, repeat the order matching decision iteration based on the reorganization scheduling decision parameters, the fitness, and the NSGA-II algorithm returned by the reorganization scheduling planning subsystem until the target order matching parameters are generated, and use the received target reorganization scheduling decision parameters and the target order matching parameters as the collaboration result. The target reorganization scheduling decision parameters are the reorganization scheduling decision parameters used to generate the target order matching parameters; 2. The method according to claim 1, wherein The reorganization scheduling decision parameters at least include the processing cost parameter per process unit, the completion time parameter, and the order online time parameter. The method further includes: Determine all the reorganization orders corresponding to the order matching parameters issued before the current upper-layer decision, and encode and initialize the population of all the reorganization orders according to the upper-layer optimization model corresponding to the preset two-layer optimization mathematical model to generate an initial encoded individual. Among them, the upper-layer optimization model includes the AEVB product difference function, the reorganization cost function, the late penalty cost function, and the line-side warehousing cost function. The reorganization cost function is associated with the processing cost parameter per process unit, the late penalty cost function is associated with the completion time parameter, and the line-side warehousing cost function is associated with the order online time parameter. The sub-encoding corresponding to the initial encoded individual is used to represent a reorganization product matching decision; Based on the NSGA-II algorithm, the upper-layer optimization model, and the fitness corresponding to all the initial encoded individuals, perform non-dominated sorting of the NSGA-II algorithm on the multiple initial encoded individuals, and perform genetic evolution operations on the candidate encoded individuals that have completed non-dominated sorting to generate the current encoded individual corresponding to the current upper-layer decision. Among them, the fitness is determined according to the sub-individual objective value calculated by coupling the AEVB product difference function, the reorganization cost function, the late penalty cost function, and the line-side warehousing cost function for the sub-encoding corresponding to the initial encoded individual. The genetic evolution operations include tournament selection operation, multi-point crossover operation, and non-uniform mutation operation; Use the order matching parameters generated by decoding the current encoded individual as the updated order matching parameters.

3. The method according to claim 2, wherein Perform genetic evolution operations on the candidate coded individuals that have completed non-dominated solution sorting, including: Randomly select any two of the multiple alternative coded individuals obtained from the tournament selection operation on the multiple candidate coded individuals as an intended coded individual group, and after selecting multiple sub-codes to be cross-operated from all the sub-codes corresponding to the two alternative coded individuals in the intended coded individual group, cross the sub-codes selected from the corresponding two alternative coded individuals to generate two corresponding first-generation offspring coded individuals; Select the sub-codes to be mutated from all the sub-codes of the two corresponding first-generation offspring coded individuals in each group of intended offspring coded individual groups, and determine at least one intended code among the selected sub-codes, where the intended code is used to represent a code with a non-zero corresponding coded value; Randomly increase or decrease the at least one intended code according to a preset mutation intensity control number, and use the mutated code as the new code of the corresponding sub-code to generate a second-generation offspring coded individual corresponding to each of the first sub-coded individuals, and obtain the coded individual corresponding to the current upper-level decision for the current time.

4. The method according to claim 3, characterized in that, Before generating the target order matching parameters, the method further includes: determining whether the order matching decision iteration satisfies a preset iteration termination condition after the current upper-level decision, where the iteration termination condition includes at least one of the following: the fitness difference between the current upper-level decision iteration and the coded individual corresponding to the previous upper-level decision iteration is less than the population fitness change threshold, and the number of times of the order matching decision iteration exceeds the preset iteration number threshold; When it is determined that the iteration termination condition is satisfied, use the order matching parameters generated by the current upper-level decision as the target order matching parameters; When it is determined that the iteration termination condition is not satisfied, perform the next order matching decision iteration based on the recombination scheduling decision parameters, the fitness, and the NSGA-II algorithm returned by the recombination scheduling planning subsystem.

5. The method according to claim 1, wherein The recombination scheduling planning subsystem makes recombination scheduling decisions based on the improved adaptive catastrophic genetic algorithm and the corresponding order matching parameters, including: After receiving the corresponding order matching parameters, the recombination scheduling planning subsystem obtains all the recombination orders from the corresponding order matching parameters and determines the target operation process parameters specified for each recombination order; The recombination scheduling planning subsystem performs integer coding and initialization on the recombination orders and the target operation process parameters according to a preset coding rule based on the lower-level optimization model corresponding to the double-layer optimization mathematical model to generate multiple recombination scheduling coded individuals, where the lower-level optimization model includes a completion time objective function and multiple target sub-constraint parameters, and the sub-codes corresponding to the recombination scheduling coded individuals are encoded based on the completion time objective function and the target sub-constraint parameters and are used to represent the sorting of the recombination orders associated with one target operation process parameter; The described recombination scheduling planning subsystem performs iterative genetic evolution operations and population catastrophe operations on multiple corresponding recombination scheduling code bodies based on a preset catastrophe adaptive genetic algorithm, the lower-layer optimization model, and the scheduling fitness corresponding to the recombination scheduling code body participating in each recombination scheduling iteration, until multiple intended recombination scheduling code bodies are generated. Among them, the scheduling fitness is calculated and determined according to the completion time objective function corresponding to the sub-code corresponding to the recombination scheduling code body; The recombination scheduling planning subsystem obtains a target recombination scheduling code body from multiple intended recombination scheduling code bodies, and after decoding the target recombination scheduling code body, generates the recombination scheduling decision parameters corresponding to the current lower-layer decision.

6. The method according to claim 5, wherein The iterative genetic evolution operations and population catastrophe operations performed by the recombination scheduling planning subsystem on multiple corresponding recombination scheduling code bodies include: The recombination scheduling planning subsystem selects a preset number of alternative scheduling code bodies from the multiple recombination scheduling code bodies participating in the current recombination scheduling iteration based on the fitness of the multiple recombination scheduling code bodies participating in the current recombination scheduling iteration and using an elitist retention strategy; The recombination scheduling planning subsystem performs partial match crossover operations and insertion mutation operations on all the alternative scheduling code bodies in sequence based on corresponding adaptive crossover probabilities and adaptive mutation probabilities to generate multiple first coding individuals. Among them, the recombination scheduling code body corresponding to the current genetic evolution operation iteration includes the first coding individuals. The adaptive crossover probability is determined based on a crossover probability function corresponding to the catastrophe adaptive genetic algorithm, and the adaptive mutation probability is determined based on a mutation probability function corresponding to the catastrophe adaptive genetic algorithm; The recombination scheduling planning subsystem selects the recombination scheduling code body with the optimal fitness from all the recombination scheduling code bodies generated in the previous genetic evolution iteration to obtain the first optimal individual corresponding to the previous genetic evolution iteration, and selects the first coding individual with the optimal fitness from the multiple first coding individuals to obtain the second optimal individual corresponding to the current genetic evolution iteration, and determines whether the fitness of the second optimal individual is higher than that of the first optimal individual; The recombination scheduling planning subsystem selects the target optimal individual corresponding to the current genetic evolution iteration from the second optimal individual and the first optimal individual according to the judgment result, and performs a catastrophe operation on the multiple first coding individuals in a preset catastrophe manner to generate multiple second coding individuals. Among them, the recombination scheduling code body corresponding to the current population catastrophe operation iteration includes the second coding individuals.

7. The method according to claim 6, wherein The recombination scheduling planning subsystem performs a catastrophe operation on the multiple first coding individuals in a preset catastrophe manner to generate multiple second coding individuals, including: When the recombinant scheduling planning subsystem determines that the fitness of the second optimal individual is higher than that of the first optimal individual, it selects a preset number of the first encoded individuals in the order of decreasing fitness to obtain elite encoded individuals, and initializes all the elite encoded individuals based on the lower-layer optimization model to generate corresponding multiple second encoded individuals; When the recombinant scheduling planning subsystem determines that the fitness of the second optimal individual is not higher than that of the first optimal individual, it selects the number of first encoded individuals equal to the set catastrophe scale in the order of increasing fitness to obtain eliminated encoded individuals, and after removing all the eliminated encoded individuals from the multiple first encoded individuals, adds a plurality of newly generated encoded individuals randomly generated to the remaining first encoded individuals to obtain corresponding multiple second encoded individuals, where the number of the newly generated encoded individuals is equal to the number of the set catastrophe scale.

8. The method according to claim 7, wherein After obtaining corresponding multiple second encoded individuals, the method further includes: The recombinant scheduling planning subsystem determines whether the current iteration of the population catastrophe operation satisfies a preset iteration termination condition after the current iteration of the population catastrophe operation, where the iteration termination condition includes at least one of the following: the fitness difference between the second encoded individuals corresponding to the current iteration of the population catastrophe operation and the previous iteration of the population catastrophe operation is less than the population fitness change threshold, and the number of iterations of the current iteration of the population catastrophe operation exceeds the preset iteration number threshold; When the recombinant scheduling planning subsystem determines that the iteration termination condition is satisfied, it uses the multiple second encoded individuals as multiple intended recombinant scheduling encoded bodies; When the recombinant scheduling planning subsystem determines that the iteration termination condition is not satisfied, it performs iterations of genetic evolution operations and population catastrophe operations on the multiple second encoded individuals until multiple intended recombinant scheduling encoded bodies are obtained.

9. The method according to claim 5, characterized in that Before generating multiple first encoded individuals, the method further includes: The recombinant scheduling planning subsystem determines the average fitness and the minimum fitness corresponding to all the alternative scheduling encoded bodies according to the fitness of the alternative scheduling encoded bodies, and determines whether the average fitness is less than the minimum fitness; When the recombinant scheduling planning subsystem determines that the average fitness is less than the minimum fitness, it uses the preset maximum crossover probability and maximum mutation probability as the corresponding adaptive crossover probability and adaptive mutation probability respectively; When the recombinant scheduling planning subsystem determines that the average fitness is greater than the minimum fitness, it decreases the maximum crossover probability and the maximum mutation probability respectively by using the corresponding crossover probability function and mutation probability function, and uses the first crossover probability and the first mutation probability obtained after the decrease as the corresponding adaptive crossover probability and adaptive mutation probability respectively.

10. The method according to claim 5, wherein The recombinant scheduling planning subsystem obtains a target recombinant scheduling encoded body from multiple intended recombinant scheduling encoded bodies, including: The recombinant scheduling planning subsystem determines the fitness corresponding to each of the intended recombinant scheduling coding bodies; The recombinant scheduling planning subsystem selects the intended recombinant scheduling coding body with the highest fitness from multiple intended recombinant scheduling coding bodies to obtain the target recombinant scheduling coding body, and extracts all sub-codings of the target recombinant scheduling coding body to obtain target sub-codings; The recombinant scheduling planning subsystem decodes the recombinant scheduling decision information corresponding to all the target sub-codings, and uses the recombinant decision information corresponding to all the target sub-codings as the recombinant scheduling decision parameters returned in response to the corresponding order matching parameters.

Images