Resource allocation method, device, equipment and medium for photovoltaic module production

Abstract

The embodiment of the present application provides a resource allocation method, device, computer equipment and readable storage medium for photovoltaic module production, the method comprising: obtaining order information of each order within the production scheduling cycle, the order information including the required power gear of the module; obtaining the available resource range of each order according to the order information; based on the available resource range of each order, multi-objective optimization is performed on multiple pre-constructed objective functions to obtain a target resource allocation strategy, the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle; the multiple objective functions include a first objective function for calculating the downshift rate and a second objective function for calculating the by-product rate. The embodiment of the present application can obtain a resource allocation scheme that is close to the global optimal solution through a multi-objective optimization algorithm, and intelligently perform a better photovoltaic module production resource allocation, reduce resource waste and improve production efficiency.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of production resource configuration, and in particular, to a resource configuration method, apparatus, computer equipment, and computer-readable storage medium for photovoltaic module production. Background Art

[0002] Photovoltaic modules, also known as solar cell modules, are devices used to convert sunlight into electrical energy. The core component of a photovoltaic module is a solar cell. A certain number of cells are connected together to form a photovoltaic module. In the actual photovoltaic market, product demand is diverse, such as demand for photovoltaic modules of different quantities and / or different powers. The diversity of product demand leads to a high degree of uncontrollability of resource allocation in the production process. For example, currently, resource allocation is based on manual experience, which often leads to misaligned resource allocation and serious waste.

[0003] It should be noted that the above content is not necessarily prior art, nor is it intended to limit the scope of patent protection of this application. Summary of the invention

[0004] The embodiments of the present application provide a resource allocation method, apparatus, computer equipment, and computer-readable storage medium for photovoltaic module production to solve or alleviate one or more of the technical problems raised above.

[0005] One aspect of an embodiment of the present application provides a resource allocation method for photovoltaic module production, the method comprising:

[0006] Obtaining order information of each order within a production scheduling cycle, wherein the order information includes a required power level of a component;

[0007] According to the order information, obtain the available resource range of each order;

[0008] Based on the available resource range of each order, a plurality of pre-built objective functions are optimized to obtain a target resource allocation strategy, wherein the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle;

[0009] The multiple objective functions include a first objective function for calculating a downshift rate and a second objective function for calculating a byproduct rate, wherein the downshift rate represents a proportion of components higher than the required power gear, and the byproduct rate represents a proportion of components lower than the required power gear.

[0010] Optionally, the available resource range includes a material combination range and a battery cell gear range.

[0011] Optionally, the order information further includes material combination restrictions and battery cell position restrictions;

[0012] Correspondingly, obtain the available resource range for each order, including:

[0013] Obtaining the supply of various materials within the production scheduling cycle and the supply of battery cells at each battery cell level;

[0014] According to the material combination restrictions of each order, the battery cell position restrictions, the supply of various materials within the production scheduling cycle and the battery cell supply of each battery cell position, the material combination range and battery cell position range of each order are obtained.

[0015] Optionally, the multi-objective optimization of the pre-built multiple objective functions based on the available resource range of each order to obtain the target resource allocation strategy includes:

[0016] Generate an initial resource allocation strategy for each order based on the available resource range of each order;

[0017] Based on the initial resource allocation strategy and preset constraints of each order, the target resource allocation strategy is obtained.

[0018] Optionally, the generation operation of the initial resource allocation strategy for each order includes:

[0019] Within the available resource range of the order, a first genetic algorithm is used to generate and iterate N initial populations, each of which includes multiple individuals, each of which includes a material combination and a group of battery cell gear ratios, where N is an integer greater than 1;

[0020] Performing genetic processing on the Nth initial population to obtain a target offspring population corresponding to the Nth initial population;

[0021] Obtaining a target individual in the target offspring population, wherein the target individual is used to generate an initial resource allocation strategy for an order;

[0022] The first genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or a mutation operation.

[0023] Optionally, N initial populations are generated and iterated, including:

[0024] Generate the first initial population within the available resources of the order;

[0025] Obtaining the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions;

[0026] The second initial population and subsequent initial populations are iterated in sequence: each initial population is obtained by genetically processing the previous initial population based on the downshift rate and byproduct rate of each individual in the previous initial population.

[0027] Optionally, a second initial population and subsequent initial populations are iterated in sequence, including:

[0028] In the process of obtaining one of the target initial populations, a crossover and / or mutation operation is performed on a plurality of individuals in an initial population previous to the target initial population, so as to obtain a descendant population corresponding to the initial population previous to the target initial population;

[0029] Merging the previous initial population of the target initial population and the corresponding offspring population to obtain a merged population;

[0030] Obtaining the downshift rate and byproduct rate corresponding to each individual in the merged population through the multiple objective functions;

[0031] According to the downshift rate and by-product rate corresponding to each individual, some individuals are selected from the merged population to obtain the target initial population.

[0032] Optionally, selecting some individuals from the merged population according to the downshift rate and byproduct rate corresponding to each individual includes:

[0033] According to the downshift rate and the byproduct rate corresponding to each individual, the merged population is divided into a plurality of non-dominated layers, each non-dominated layer corresponds to a non-dominated sequence value;

[0034] According to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence.

[0035] Optionally, according to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence, including:

[0036] According to the non-dominated sequence values ​​of the multiple non-dominated layers, some individuals of each non-dominated layer are obtained in sequence to obtain multiple individuals;

[0037] When the number of the plurality of individuals does not satisfy the preset population size, one or more individuals having the minimum vertical distance to the preset reference point are continuously obtained from the non-dominated layer with the minimum non-dominated sequence value.

[0038] Optionally, according to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence, including:

[0039] According to the non-dominated sequence values ​​of the multiple non-dominated layers, multiple individuals satisfying the preset population size are preferentially obtained from the multiple non-dominated layers with small non-dominated sequence values;

[0040] In the last non-dominated layer providing the plurality of individuals, an expected number of individuals are obtained according to the individual crowding distance.

[0041] Optionally, the order information further includes an order demand and a benchmark power table, wherein the order demand includes the demand for each power level of the component of the order, each power level includes the required power level, and the benchmark power table includes different benchmark powers of the component under different material combinations, wherein one material combination corresponds to one benchmark power;

[0042] Correspondingly, the downshift rate and byproduct rate of each individual in the first initial population are obtained, including:

[0043] According to the individual material combination and the reference power table, obtaining the corresponding reference power;

[0044] Obtaining a component power distribution matrix according to the corresponding reference power and the order demand;

[0045] Obtain the demand-output gap of the component based on the individual material combination, the battery cell gear ratio and the component power distribution matrix;

[0046] According to the demand-output gap, the downshift rate and by-product rate corresponding to the individual are obtained.

[0047] Optionally, based on the initial resource allocation strategy and preset constraints of each order, the target resource allocation strategy is obtained, including:

[0048] The first global initial population is constructed as a global individual using the initial resource allocation strategy of each order, and the second global initial population is iterated sequentially to the Mth global initial population under the preset constraints using the second genetic algorithm; wherein M is an integer greater than 2;

[0049] Based on the preset constraint conditions, performing genetic processing on the Mth global initial population to obtain a target global offspring population corresponding to the Mth global initial population;

[0050] Acquire a target global individual in the target global offspring population, wherein the target global individual is used to generate the target resource allocation strategy;

[0051] The second genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or mutation operation.

[0052] Optionally, a first global initial population is constructed as a global individual using the initial resource allocation strategy of each order, and a second global initial population is iterated sequentially to an Mth global initial population under the preset constraints using a second genetic algorithm using the first global initial population as a starting point, including:

[0053] Take the initial resource allocation strategy of each order as the global individual and construct the first global initial population;

[0054] According to the preset constraint condition, some global individuals in the first global initial population are eliminated;

[0055] Obtaining the downshift rate and byproduct rate of the remaining global individuals in the first global initial population through the multiple objective functions;

[0056] The second global initial population and the subsequent global initial populations are iterated in sequence: each global initial population is obtained by genetic processing based on the downshift rate and byproduct rate of the remaining global individuals in the previous global initial population.

[0057] Optionally, the order information further includes order scheduling quantity, and the preset constraint condition includes the gap range between supply and demand of battery cells;

[0058] Correspondingly, according to the preset constraint condition, some global individuals in the first global initial population are eliminated, including:

[0059] Obtaining the battery cell supply ratio of each battery cell gear within the production scheduling cycle;

[0060] According to the order scheduling volume of each order and the battery cell gear ratio of each global individual in the first global initial population, respectively obtain the battery cell demand ratio of each global individual in each battery cell gear;

[0061] According to the battery cell supply ratio and the battery cell demand ratio, obtaining the battery cell supply and demand gap corresponding to each global individual;

[0062] According to the battery supply-demand gap corresponding to each global individual and the battery supply-demand gap range, some global individuals in the first global initial population are eliminated.

[0063] Another aspect of the embodiments of the present application provides a resource allocation device for photovoltaic module production, the device comprising:

[0064] A first acquisition module is used to acquire order information of each order in a production scheduling cycle, wherein the order information includes a required power level of a component;

[0065] A second acquisition module, used to acquire the available resource range of each order according to the order information;

[0066] The optimization module is used to perform multi-objective optimization on multiple pre-built objective functions based on the available resource range of each order to obtain the target resource allocation strategy;

[0067] Among them, the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle, and the multiple objective functions include a first objective function for calculating the downshift rate and a second objective function for calculating the by-product rate. The downshift rate represents the proportion of components higher than the required power level, and the by-product rate represents the proportion of components lower than the required power level.

[0068] Another aspect of an embodiment of the present application provides a computer device, including:

[0069] at least one processor; and

[0070] a memory communicatively coupled to the at least one processor;

[0071] Wherein: the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0072] Another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method described above is implemented.

[0073] The above technical solution adopted in the embodiment of the present application may have the following advantages:

[0074] By obtaining the order information of each order within the production scheduling cycle, the available resource range of each order is determined. Within the available resource range of each order, a target resource allocation strategy that achieves the best balance between the downshift rate and the by-product rate is found through a multi-objective optimization algorithm. The target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle. It can be seen that the embodiment of the present application can obtain a resource allocation plan that is close to the global optimal solution through a multi-objective optimization algorithm, and intelligently perform a better photovoltaic module production resource allocation, reduce resource waste and improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0076] Figure 1A flowchart of a resource configuration method for photovoltaic module production according to Embodiment 1 of the present application is schematically shown;

[0077] Figure 2 Schematically shows Figure 1 Flow chart of sub-steps of step S102;

[0078] Figure 3 Schematically shows Figure 1 Flow chart of sub-steps of step S104;

[0079] Figure 4 Schematically shows Figure 3 Flow chart of sub-steps of step S300;

[0080] Figure 5 Schematically shows Figure 4 Flow chart of sub-steps of step S400;

[0081] Figure 6 Schematically shows Figure 5 Flow chart of sub-steps of step S504;

[0082] Figure 7 Schematically shows Figure 6 Flow chart of sub-steps of step S606;

[0083] Figure 8 Schematically shows Figure 3 Flow chart of sub-steps of step S302;

[0084] Fig. 9 is an application example diagram of the resource configuration method for photovoltaic module production according to the first embodiment of the present application;

[0085] Fig.10 is another application example diagram of the resource allocation method for photovoltaic module production according to the first embodiment of the present application;

[0086] Fig.11 The flowchart of the genetic algorithm of the first embodiment of the present application is schematically shown;

[0087] Fig.12 A block diagram schematically shows a resource allocation device for photovoltaic module production according to the second embodiment of the present application;

[0088] Fig.13 The hardware architecture diagram of the computer device according to the third embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0089] In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail 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 intended to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0090] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0091] In the description of the present application, it should be understood that the numerical labels before the steps do not indicate the order in which the steps are executed, but are only used to facilitate the description of the present application and to distinguish each step, and therefore should not be understood as a limitation on the present application.

[0092] First, the following terms are explained:

[0093] Photovoltaic module: also known as solar cell module, is one of the main components of photovoltaic system. Photovoltaic module is used to convert sunlight into electrical energy, and its core component is solar cell. Multiple solar cells can form a complete photovoltaic module by connecting and assembling.

[0094] Multi-objective optimization algorithm: used to solve multi-objective optimization problems. In a multi-objective optimization problem, there can be multiple objective functions, each of which represents an optimization direction of the problem, and there can be conflicts between multiple objective functions. The goal of the algorithm is to find the Pareto optimal solution (Pareto), which is a set of solutions that cannot be dominated (defeated) by other solutions on all objective functions, and they together constitute the Pareto front. The Pareto front is a specific subset of the problem that includes all Pareto solutions.

[0095] Battery cell: in the embodiments of the present application, it refers to a solar cell, but it may also be other battery cells (such as battery cells for new energy vehicles).

[0096] Cell level: used to reflect the efficiency of the cell, that is, the ability of the cell to convert sunlight into electrical energy. Different cells can have different efficiency levels, corresponding to different cell levels. Cell levels can be expressed using numbers, letters or efficiency percentages, for example: "Grade A", "Grade 3", "22%", etc.

[0097] Battery cell level ratio: the proportion of battery cells used in each battery cell level.

[0098] Battery supply and demand gap: A value that describes the degree of balance between the predicted battery cell gear ratio in actual production scheduling. Too low a value will lead to production delays, and too high a value will lead to waste of resources.

[0099] Module power range: used to reflect the ability of photovoltaic modules to convert solar energy into electrical energy. Different photovoltaic modules can have different powers, corresponding to different module power ranges. For example: 5W (watt) one range, 595W represents the module power range of [595, 600), and 600W represents the module power range of [600, 605).

[0100] Iverson bracket: A mathematical notation used to simplify and compactly describe mathematical expressions and logical operations. Iverson brackets are usually represented by square brackets [], which contain a conditional expression. If the condition is met, the value of the Iverson bracket is 1, and if the condition is not met, the value of the Iverson bracket is 0.

[0101] Niche distance: In genetic algorithms, it is used to measure the similarity or distance between two individuals. A smaller niche distance indicates that the two individuals are similar in the search space.

[0102] Secondly, in order to facilitate those skilled in the art to understand the technical solutions provided in the embodiments of the present application, the relevant technologies are described below:

[0103] In the production and use of photovoltaics, product demands are diverse, such as demands for different quantities and / or different powers of photovoltaic modules. The above diversity relies on customized production, and each customized product (such as photovoltaic modules) must meet a specific power requirement. The above diversity leads to a high degree of uncontrollability of resource allocation in the production process. For example, currently, all configurations rely on manual experience, which often leads to misaligned resource allocation and serious waste.

[0104] Specifically, the applicant understands that the photovoltaic manufacturing field mainly relies on the experience of business-related personnel to arrange resource allocation. However, there are many deficiencies in resource allocation: in the face of large-volume orders and a sharp expansion of decision-making space, it is difficult for manual experience to ensure that the production plan can meet the supply and demand balance of photovoltaic modules to the greatest extent while maintaining the supply and demand balance of battery cells. In addition, the products or production processes of different workshops and production lines are complex and diverse, and the number of products, raw materials and spare parts is huge. It is difficult to gain insight into the wide range of possibilities by relying solely on manual experience. When time is tight, it is difficult to provide the best arrangement from a global perspective, resulting in misallocation and serious waste of production resources.

[0105] To this end, the embodiment of the present application provides a resource allocation technology solution for photovoltaic module production. In this technical solution: (1) data preprocessing, business simulation, and multi-objective optimization models are set up, and can adapt to different products and actual business needs, and realize high-efficiency and optimal resource allocation solutions (such as battery cell allocation) for any order; (2) a resource allocation solution that is close to the global optimal solution is used to achieve more refined production resource allocation, less resource mismatch and resource waste, and realize order delivery with maximum production efficiency and resource conservation; (3) the supply and demand balance of photovoltaic modules can be met to the greatest extent while maintaining the supply and demand balance of battery cells. See below for details.

[0106] The technical solutions of the present application are described below through multiple embodiments. It should be noted that these embodiments can be implemented in a variety of different forms and should not be construed as being limited to the embodiments described here.

[0107] Embodiment 1

[0108] Figure 1 The flowchart of the resource configuration method for photovoltaic module production according to the first embodiment of the present application is schematically shown.

[0109] like Figure 1 As shown, the resource allocation method for photovoltaic module production may include steps S100 to S104, wherein:

[0110] Step S100, obtaining order information of each order in a production scheduling cycle, wherein the order information includes a required power level of a component.

[0111] Step S102: acquiring the available resource range of each order according to the order information.

[0112] Step S104, based on the available resource range of each order, multi-objective optimization is performed on the pre-constructed multiple objective functions to obtain a target resource allocation strategy, where the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle.

[0113] The multiple objective functions include a first objective function for calculating a downshift rate and a second objective function for calculating a byproduct rate, wherein the downshift rate represents a proportion of components higher than the required power gear, and the byproduct rate represents a proportion of components lower than the required power gear.

[0114] The resource allocation method for photovoltaic module production provided in this embodiment determines the available resource range of each order by obtaining the order information of each order within the production scheduling cycle. Within the available resource range of each order, a target resource allocation strategy that achieves the best balance between the downshift rate and the by-product rate is found through a multi-objective optimization algorithm, and the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle. It can be seen that the embodiment of the present application can obtain a resource allocation strategy that is close to the global optimal solution through a multi-objective optimization algorithm, thereby intelligently performing more accurate and reasonable photovoltaic module production resource allocation, reducing resource waste and improving production efficiency.

[0115] The following combination Figure 1 , each step in steps S100 to S104 and other optional steps are described in detail.

[0116] Step S100, obtaining order information of each order in a production scheduling cycle, wherein the order information includes a required power level of a component.

[0117] The production scheduling cycle is a time unit set by a factory or production unit in a production plan, usually measured in days, weeks, or months. Within a production scheduling cycle, the factory can plan to complete the production tasks of one or more orders. For example, two hundred orders need to be produced within a one-week production scheduling cycle. The orders may involve various products, such as automobiles, hinges, etc. The orders in the embodiments of the present application are for the production needs of photovoltaic modules. The order information may include detailed data of the order, such as: the required power level of the (photovoltaic) module, the required quantity of the module, etc. By obtaining the order information of each order within the production scheduling cycle, the demand for each order can be understood, which is used for resource allocation.

[0118] Step S102: acquiring the available resource range of each order according to the order information.

[0119] Resources can include various means of production, such as production equipment, testing equipment, inventory, material combinations and cells. Among them, the material combination is a collection of materials used to manufacture photovoltaic modules, which can include coating materials (such as single coating, double coating), packaging materials (such as glue), backplane materials (such as aluminum, stainless steel), etc. Different materials can be combined in different ways. Cells are the basic units of photovoltaic modules and are used to assemble photovoltaic modules. Due to the different specific requirements of the order, the available resources will also be subject to corresponding restrictions.

[0120] In this embodiment, the available resource range of each order is calculated based on the order information, so that resource allocation can be better performed.

[0121] In an optional embodiment, the available resource range includes a material combination range, a battery cell range, and other resource ranges formed according to demand. Among them, the material combination range is used to indicate one or more material combinations that can be used for the order, such as V1, V2, V3, and V4. The battery cell range is used to indicate which efficiency of batteries can be used for the order. For example, an order requires the production of ladder-type photovoltaic modules. The battery cell range of the order can be divided into levels S1 to S16. For example, battery cell level S1 can correspond to a battery cell with a power generation efficiency of "22%".

[0122] It should be noted that there are many ways to obtain the scope of available resources for an order, such as analyzing order information, evaluating existing resources, confirming supply chain conditions, coordinating related resources, etc. An exemplary acquisition scheme is provided below.

[0123] In an optional embodiment, if Figure 2 As shown, the order information also includes material combination restrictions and battery cell position restrictions. Correspondingly, step S102 may include:

[0124] Step S200, obtaining the supply quantity of various materials within the production scheduling cycle and the supply quantity of battery cells at each battery cell level.

[0125] Step S202, obtaining the material combination range and battery cell position range of each order according to the material combination restriction of each order, the battery cell position restriction, the supply of various materials in the production scheduling cycle and the battery cell supply of each battery cell position.

[0126] Based on the factory inventory, self-production capacity and supply plan, the supply of various materials and the supply of cells in each cell level within the production cycle can be calculated. Combined with the actual situation of the order, that is, the material combination restrictions and cell level restrictions extracted from the order information, the material combination range and cell level range of each order are calculated.

[0127] In this embodiment, available material combinations and cell positions are found for each order based on supply capabilities and constraints, so that resources can be better allocated, correct and effective use of resources can be ensured, and waste caused by mismatch can be reduced.

[0128] Step S104, based on the available resource range of each order, multi-objective optimization is performed on the pre-constructed multiple objective functions to obtain a target resource allocation strategy, where the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle.

[0129] The multiple objective functions can be set as multiple functions according to needs. In the present embodiment, they can include a first objective function for calculating a downshift rate and a second objective function for calculating a byproduct rate. The downshift rate represents the proportion of components that are higher than the required power gear, and the byproduct rate represents the proportion of components that are lower than the required power gear.

[0130] In the production process of photovoltaic modules, affected by multiple factors such as resource allocation, production process, production environment, and human error, the power of photovoltaic modules actually produced may not completely correspond to the required power level of the modules, and the power of a small number of photovoltaic modules may be higher or lower than the required power level. For example, if the required power level is 550W, the power of photovoltaic modules actually produced may present a normal distribution centered on 550W. Among them, the components that are higher than the required power level are called downshifts, and the components that are lower than the required power level are called byproducts. Excessive downshifts and excessive byproduct rates will lead to resource waste and increased costs. In order to alleviate the above problems, based on the available resource range of each order, multiple objective functions can be constructed with the minimization of downshift rate and byproduct rate as optimization objectives, and the target resource allocation strategy can be found through a multi-objective optimization algorithm, that is, the optimal resource allocation solution that strikes a balance between downshift rate and byproduct rate. Among them, the multi-objective optimization algorithm can include genetic algorithm, particle swarm optimization algorithm, annealing algorithm, ant colony optimization algorithm, differential evolution algorithm, etc. It should be noted that the objective function is not limited to minimizing the downshift rate and the by-product rate, but can also be constructed according to actual needs, such as minimizing costs, maximizing resource savings, maximizing resource utilization, meeting order delivery time requirements, etc.

[0131] In this embodiment, a multi-objective optimization algorithm is used to find a target resource allocation strategy that achieves the best balance between the downshift rate and the by-product rate, that is, a resource allocation solution that is close to the global optimal solution, so that better resource allocation can be implemented in the production process, reducing resource waste, improving quality and efficiency, and meeting different order requirements.

[0132] Multi-objective optimization can directly perform a single optimization to obtain the target resource allocation strategy.

[0133] In order to improve the performance of multi-objective optimization, the multi-objective optimization process may also be improved. An exemplary solution is provided below.

[0134] In an optional embodiment, if Figure 3 As shown, step S104 may include:

[0135] Step S300: generating an initial resource allocation strategy for each order based on the available resource range of each order.

[0136] Step S302: acquiring the target resource allocation strategy based on the initial resource allocation strategy and preset constraints of each order.

[0137] In this embodiment, the multi-objective optimization is divided into two stages: in the first stage, each order is optimized separately, and one or more initial resource allocation strategies corresponding to each order are obtained under the condition of minimum downshift rate and minimum by-product rate. In the second stage, all orders are optimized globally, and the initial resource allocation strategies of each order obtained in the first stage are summarized. While considering the two objectives, the algorithm convergence is accelerated and the calculation accuracy is improved by presetting constraints, and finally the target resource allocation strategy is obtained, which is the best resource allocation scheme from a global perspective. Among them, the preset constraints are used to improve the rationality and feasibility of the resource allocation strategy, and can be set according to actual needs, such as resource constraints, delivery time, supply and demand gap, etc. It can be understood that in this embodiment, multi-objective optimization is carried out in two stages, and the optimization results of the first stage are used as the initial range of the second stage, which can effectively improve the optimization accuracy and calculation efficiency.

[0138] Wherein, step S300 is to optimize each order separately based on the available resource range of each order to obtain the corresponding initial resource allocation strategy. Figure 4 As shown, step S300 may include:

[0139] Step S400: within the available resource range of the order, generate and iterate N initial populations using a first genetic algorithm, each initial population including multiple individuals, each individual including a material combination and a group of battery cell gear ratios, wherein N is an integer greater than 1.

[0140] Step S402: Perform genetic processing on the Nth initial population to obtain a target offspring population corresponding to the Nth initial population.

[0141] Step S404, obtaining a target individual in the target offspring population, wherein the target individual is used to generate an initial resource allocation strategy for an order.

[0142] The first genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or a mutation operation.

[0143] As described above, the multi-objective optimization algorithm may include various algorithms for optimization, such as a first genetic algorithm, wherein the first genetic algorithm may preferably adopt NSGA-III genetic algorithm (Non-dominated Sorting Genetic Algorithm-Ⅲ, third generation non-dominated sorting genetic algorithm), or other genetic algorithms may be adopted.

[0144] The following describes the process of obtaining the initial resource allocation strategy for an order (such as order A): Based on the material combination range and battery cell gear range of order A, N initial populations are generated and iterated using the first genetic algorithm. Each initial population includes multiple individuals, each of which includes a material combination and a set of battery cell gear ratios, wherein the battery cell gear ratio is the battery cell usage ratio of each battery cell gear, such as the usage ratio of ladder-type photovoltaic module batteries S1 to S16. Each individual represents a resource allocation strategy, that is, a potential solution to the problem. N is an integer greater than 1, and the specific value can be determined according to the preset number of iterations.

[0145] After multiple rounds of iterations, the Nth initial population for order A is obtained, and genetic processing is performed on the Nth initial population to obtain the target offspring population. Each individual in the target offspring population corresponds to an initial resource allocation strategy for order A. It should be noted that, according to the preset indicator (objective function) priority, some individuals can be obtained from the target offspring population and used as the initial resource allocation strategy for order A.

[0146] When the first genetic algorithm adopts the NSGA-III genetic algorithm, each individual in the target offspring population corresponds to a Pareto optimal solution, forming a Pareto solution set, and each Pareto optimal solution of the Pareto solution set represents an initial resource allocation strategy.

[0147] Genetic algorithms such as NSGA-III simulate natural selection and genetic mechanisms to continuously generate and iterate possible solutions to find the optimal solution or approximate solution to the problem. Genetic processing includes mutation and / or crossover operations. Crossover is to combine two individual parts together to produce new individuals for information exchange and inheritance. Mutation is to change the elements in an individual with a certain probability, introduce randomness, avoid local optimal solutions, and maintain the diversity of the population.

[0148] When applying genetic algorithms for multi-objective optimization, the performance and convergence speed of genetic algorithms can also be improved by adjusting hyperparameters. Hyperparameters are parameters that need to be manually set before executing the genetic algorithm. Hyperparameters can include population size, crossover probability, mutation probability, cutoff condition, number of iterations, etc. Among them, the appropriate population size can balance the diversity and convergence speed of the genetic algorithm. The appropriate crossover probability and mutation probability can help the algorithm escape from the local optimal solution. The appropriate cutoff condition and number of iterations can balance the performance and computational cost of the algorithm. By adjusting the hyperparameters, the genetic algorithm can better handle the problem and obtain faster convergence speed and higher performance.

[0149] In this embodiment, the first genetic algorithm is used to continuously iterate and improve individuals in the population, thereby finding an order-optimal or approximately optimal resource allocation solution, thereby achieving order-level resource configuration optimization and improving the quality of resource allocation.

[0150] In an optional embodiment, if Figure 5 As shown, step S400 "generating and iterating N initial populations using the first genetic algorithm" may include:

[0151] Step S500, generating a first initial population within the available resource range of the order.

[0152] Step S502: Obtain the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions.

[0153] Step S504, iterating the second initial population and subsequent initial populations in sequence: each initial population is obtained by genetically processing the previous initial population based on the downshift rate and byproduct rate of each individual in the previous initial population.

[0154] Let's continue to use order A as an example:

[0155] Construct the first initial population: Based on the material combination range and battery cell range of order A, randomly generate material combinations and the ratio of each battery cell ratio to construct individuals in the first initial population. Accordingly, the first initial population includes multiple individuals, each of which includes a material combination and a set of battery cell ratios.

[0156] Iterate out the second initial population: According to multiple objective functions, calculate the downshift rate and byproduct rate of each individual in the first initial population, so as to evaluate the quality of each individual, that is, their ability to solve the problem. Select individuals with better quality or randomly select some individuals to perform crossover and / or mutation operations, and iterate out the second initial population after genetic processing.

[0157] Iterate the third initial population: According to multiple objective functions, calculate the downshift rate and byproduct rate of each individual in the second initial population, so as to evaluate the quality of each individual, that is, their ability to solve the problem. Select individuals with better quality or randomly select some individuals to perform crossover and / or mutation operations, and iterate the third initial population after genetic processing.

[0158] Referring to the iterative process of the second initial population and the third initial population, and so on, according to the preset number of iterations, the following N-3 initial populations can be iterated in sequence, so as to finally iterate and obtain the Nth initial population of order A.

[0159] In this embodiment, during the order-level optimization process, the quality of each individual can be evaluated by calculating the downshift rate and by-product rate of each individual, and individuals are selected according to the quality of each individual to perform genetic processing to iterate the next initial population, which can gradually approach the optimal resource allocation plan and reduce computing time and resource consumption.

[0160] In an optional embodiment, in order to improve the quality of the second and subsequent initial populations, such as Figure 6 As shown, step S504 may include:

[0161] Step S600, in the process of obtaining one of the target initial populations, a crossover and / or mutation operation is performed on a plurality of individuals in an initial population previous to the target initial population, so as to obtain a descendant population corresponding to the initial population previous to the target initial population.

[0162] Step S602: merge the previous initial population of the target initial population and the corresponding offspring population to obtain a merged population.

[0163] Step S604: Obtain the downshift rate and byproduct rate corresponding to each individual in the merged population through the multiple objective functions.

[0164] Step S606 , selecting some individuals from the merged population according to the downshift rate and byproduct rate corresponding to each individual, so as to obtain the target initial population.

[0165] As an example, the target initial population is any one of the second to Nth initial populations. The following takes the second initial population as the target initial population to illustrate the specific process of obtaining the target initial population: The first initial population P is randomly generated based on the available range of the order. t , the population size is J. t Obtain the offspring population Q corresponding to the first initial population through crossover and mutation t , the offspring population Q t The population size can be J. The initial population P t and the offspring population Q t Combined, we get a new merged population R t , the population size is 2J. Calculate the combined population R t The downshift rate and byproduct rate of each individual in the , and then evaluate the quality (fitness) of each individual. According to the quality (fitness) of each individual, select J individuals from the merged population to construct a new population S with a population size of J t , and use it as the next initial population, that is, the second initial population.

[0166] By merging the last initial population and the corresponding offspring population, the diversity of the population can be guaranteed and the performance of the genetic algorithm can be improved. Specifically in the application of this embodiment, individuals are selected according to the individual downshift rate and byproduct rate to construct the next initial population, which can maintain and enhance the quality of the population and gradually improve the quality of the resource allocation plan.

[0167] In an optional embodiment, if Figure 7 As shown, step S606 may include:

[0168] Step S700 : dividing the merged population into a plurality of non-dominated layers according to the downshift rate and the by-product rate corresponding to each individual, each non-dominated layer corresponding to a non-dominated ordinal value.

[0169] Step S702 : acquiring individuals for constructing the target initial population from a plurality of non-dominated layers in order according to the non-dominated sequence values.

[0170] According to the individual downshift rate and byproduct rate, the merged population (such as R t ) are divided into multiple non-dominated layers (F1, F2...F l ), each non-dominated layer is associated with a non-dominated sequence value, which is used to identify the position and importance of the non-dominated layer in the non-dominated sorting. The first layer F1 (the front) includes individuals that are not dominated (defeated) by solutions (individuals) in other layers under at least one objective function, and the second layer F2 includes individuals that are not dominated by solutions in the first layer under at least one objective function. These solutions may still have good performance under certain objective functions, but are worse than solutions in the first layer under other objective functions. Similarly, solutions in subsequent non-dominated layers may perform worse under certain objective functions. Therefore, individuals with lower downshift rates and by-product rates will be assigned to earlier non-dominated layers. According to the non-dominated sequence values ​​of multiple non-dominated layers, individuals used to construct the target initial population are obtained from multiple non-dominated layers in sequence.

[0171] In this embodiment, during the order-level optimization process, the merged population can be divided into multiple non-dominated layers according to the downshift rate and by-product rate of the individuals, and then individuals are selected from the multiple non-dominated layers to form the target initial population, which helps to determine the adaptability and importance of the individuals so that the optimal resource allocation plan can be found.

[0172] In an optional embodiment, step S702 may include:

[0173] According to the non-dominated sequence values ​​of the multiple non-dominated layers, some individuals of each non-dominated layer are sequentially obtained to obtain multiple individuals. When the number of the multiple individuals does not meet the preset population size, one or more individuals having the smallest vertical distance to the preset reference point are continuously obtained from the non-dominated layer with the smallest non-dominated sequence value.

[0174] For example, when merging population R t Divided into different non-dominated layers (F1, F2...F l ), select individuals from F1, then select individuals from F2, and so on, sequentially select individuals from multiple non-dominated layers to construct the target initial population S t , until S t The population size is larger than the preset population size (such as J). t From each non-dominated layer (F1, F2...F l ), but the selected individuals still do not meet the preset population size, some individuals in the critical layer (i.e., F1 layer) are accepted again. For example, according to the minimum vertical distance associated with each individual in the F1 layer and any preset reference point on the hyperplane, select K=J-|S from the F1 layer t In some embodiments, the probability of selecting an individual with a smaller fitness can be reduced according to the Hamming distance between individuals, when the Hamming distance between individuals is less than a pre-specified value (preset microhabitat distance).

[0175] In this embodiment, individuals are selected to construct a target initial population through a non-dominated sorting method of reference points, which can effectively reduce the computational cost and accelerate the convergence speed.

[0176] In another optional embodiment, step S702 may include:

[0177] According to the non-dominated sequence values ​​of the multiple non-dominated layers, multiple individuals satisfying a preset population size are preferentially obtained from the multiple non-dominated layers with small non-dominated sequence values.

[0178] In the last non-dominated layer providing the plurality of individuals, an expected number of individuals are obtained according to the individual crowding distance.

[0179] For example, the individuals of the entire layer can be obtained from the non-dominated layer with a small non-dominated sequence value, and the non-dominated layer with a small non-dominated sequence value can be sorted from low to high, until multiple individuals that meet the preset population size J are obtained. j , 1≤j≤l), individuals are obtained in descending order according to the individual crowding distance (microhabitat distance) until the preset population size is met. In some embodiments, when the combined population R t Before selecting individuals, you can first eliminate the merged population R t The check mark is an infeasible individual to improve the computational efficiency and algorithm stability. The check mark can be determined according to the preset constraints.

[0180] In this embodiment, the elite strategy and crowding operator are used to retain the excellent individuals in the parent generation and directly enter the child generation, thereby preventing the loss of the obtained Pareto optimal solution and improving the accuracy of optimization.

[0181] The above is an exemplary introduction to how to generate and iterate the initial population during the order-level optimization process.

[0182] In the iterative process, individual selection is involved, wherein the individual selection may depend on various factors including downshift rate and byproduct rate. An exemplary scheme for calculating the downshift rate and byproduct rate of an individual is provided below.

[0183] In an optional embodiment, the order information also includes an order demand and a benchmark power table, the order demand includes the demand for the order at each power level of the component, the each power level includes the required power level, and the benchmark power table includes different benchmark powers of the component under different material combinations, and one material combination corresponds to one benchmark power. Correspondingly, step S502 "obtaining the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions" can be implemented by the following process:

[0184] According to the individual material combination and the reference power table, the corresponding reference power is obtained.

[0185] A component power distribution matrix is ​​obtained according to the corresponding reference power and the order demand.

[0186] According to the individual material combination, battery cell gear ratio and the component power distribution matrix, the demand-output gap of the component is obtained.

[0187] According to the demand-output gap, the downshift rate and by-product rate corresponding to the individual are obtained.

[0188] The reference power table includes the normal distribution matrix (reference power Φ(z)) when the components use different material combinations. One material combination corresponds to one reference power Φ(z), which can be expressed as the following formula:

[0189] Φ(z) can be obtained by numerical integration or by querying a reference power table.

[0190] The order demand includes the demand for each power level (including the required power level) of the component. According to the order demand and the benchmark power table, the component power distribution matrix can be obtained, as shown below:

[0191]

[0192] Among them, pi represents the battery cell position, Match with battery cell position and materials Wm represents the power level of the module, such as 5W level, 595W represents the module power range of [595, 600). wm The demand for this order at each power level of the component.

[0193] in, It can be calculated by the following formula:

[0194]

[0195] According to the row vector corresponding to the individual (the vector elements are the material combination and the battery cell gear ratio) and the component power distribution matrix, the demand-output gap sub(x j ).

[0196] Through multiple objective functions and component demand-output gaps, the corresponding downshift rate and by-product rate of individuals are obtained.

[0197] The multi-objective optimization problem can be defined as:

[0198] in, are the downshift rate and by-product rate respectively, and F(x) is the mapping function.

[0199] Correspondingly, the first objective function for calculating the downshift rate can be expressed as the following formula:

[0200]

[0201] Among them, [] is an Iverson bracket. When the logical judgment in the bracket is true, the value of the bracket is 1, otherwise it is 0; the downshift rate is the sum of the non-negative values ​​of the demand-output gap.

[0202] Correspondingly, the second objective function for calculating the by-product rate can be expressed as the following formula:

[0203]

[0204] Among them, the by-product rate is the absolute value of the demand-output gap.

[0205] In this embodiment, by calculating the downshift rate and byproduct rate of the individual, it is helpful to find the individual corresponding to the best resource allocation strategy, thereby improving the efficiency and accuracy of the algorithm.

[0206] The above describes the multi-objective optimization of the order level in the first stage through multiple embodiments. The multi-objective optimization in the second stage will be exemplarily introduced below.

[0207] In an optional embodiment, if Figure 8 As shown, step S302 may include:

[0208] Step S800, constructing a first global initial population as a global individual based on the initial resource allocation strategy of each order, and taking the first global initial population as the starting point, using a second genetic algorithm to iterate sequentially under the preset constraints to obtain a second global initial population to the Mth global initial population; wherein M is an integer greater than 2.

[0209] Step S802: Based on the preset constraint condition, perform genetic processing on the Mth global initial population to obtain a target global offspring population corresponding to the Mth global initial population.

[0210] Step S804: obtaining a target global individual in the target global offspring population, wherein the target global individual is used to generate the target resource configuration strategy.

[0211] The second genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or mutation operation.

[0212] In the optimization process of the second stage, the initial resource allocation strategy (such as the Pareto optimal solution) of each order obtained in the first stage is used as multiple global individuals to jointly construct the first global initial population. The first global initial population is used as the starting point of the second stage multi-objective optimization, and the second global initial population to the Mth global initial population are iterated in sequence under the preset constraints by the second genetic algorithm. Among them, the second genetic algorithm and the first genetic algorithm can be the same genetic algorithm, or different genetic algorithms selected according to actual needs. M is an integer greater than 2, and the specific value can be determined according to the preset number of iterations.

[0213] After multiple rounds of iterations, the Mth initial population for the global is obtained, and genetic processing is performed on the Mth initial population to obtain the target global offspring population. Each global individual in the target global offspring population corresponds to a target resource allocation strategy in the second stage. It should be noted that one or more global individuals can also be obtained from the target global offspring population according to the preset indicator (objective function) priority, and the one or more global individuals are output to the front-end page as one or more target resource allocation strategies for users to view and / or select, so as to instruct / control the component production of the factory or workshop through the selected target resource allocation strategy.

[0214] When the second genetic algorithm also uses the NSGA-III genetic algorithm, each global individual in the target global offspring population corresponds to a second-stage Pareto optimal solution, forming a second-stage Pareto solution set, and each Pareto optimal solution of the second-stage Pareto solution set represents a target resource allocation strategy. Under the guidance of the target resource allocation strategy, the global optimal resource allocation within the production scheduling cycle can be achieved, reducing resource mismatch and resource waste.

[0215] In this embodiment, based on the initial resource allocation strategy of each order in the first stage (such as all Pareto optimal solutions for each order), the second genetic algorithm continuously iterates and improves the individuals in the population under preset constraints to quickly find the optimal resource allocation plan within the production scheduling cycle, accelerate algorithm convergence, and reduce resource waste.

[0216] In an optional embodiment, step S800 may include:

[0217] Taking the initial resource allocation strategy of each order as the global individual, the first global initial population is constructed.

[0218] According to the preset constraint condition, some global individuals in the first global initial population are eliminated.

[0219] The downshift rate and byproduct rate of the remaining global individuals in the first global initial population are obtained through the multiple objective functions.

[0220] The second global initial population and the subsequent global initial populations are iterated in sequence: each global initial population is obtained by genetic processing based on the downshift rate and byproduct rate of the remaining global individuals in the previous global initial population.

[0221] For example:

[0222] Construct the first global initial population: Use the optimal solution of all initial resource allocation strategies (such as all Pareto optimal solutions) in the first stage to construct the first global initial population.

[0223] Iterate the second global initial population: According to the preset constraints, remove some global individuals in the first global initial population that do not meet the conditions, which can reduce the unreasonable waste of computing resources, improve computing accuracy and accelerate convergence. According to multiple objective functions, calculate the downshift rate and byproduct rate of the remaining global individuals in the first global initial population, and then evaluate the quality of the global individuals, that is, their ability to solve the problem. Select global individuals with better quality or randomly select some global individuals to perform crossover and / or mutation operations, and iterate the second global initial population after genetic processing.

[0224] Iterate to generate the third global initial population: according to the preset constraints, remove some global individuals that do not meet the conditions in the second global initial population. According to multiple objective functions, calculate the downshift rate and byproduct rate of the remaining global individuals in the second global initial population, so as to evaluate the quality of the remaining global individuals, that is, their ability to solve the problem. Select global individuals with better quality or randomly select some global individuals to perform crossover and / or mutation operations, and iterate to generate the third global initial population after genetic processing.

[0225] Referring to the iterative process of the second global initial population and the third global initial population, and so on, according to the preset number of iterations, the following M-3 global initial populations can be iterated in sequence, so as to finally iterate and obtain the Mth global initial population.

[0226] In this embodiment, during the optimization process of the second stage, global individuals that do not meet the conditions are eliminated by presetting constraints, reducing infeasible solutions, effectively improving calculation accuracy and accelerating algorithm convergence speed, and further optimizing resource allocation.

[0227] In an optional embodiment, the order information further includes order scheduling, and the preset constraint condition includes a gap range between supply and demand of battery cells; correspondingly, according to the preset constraint condition, some global individuals in the first global initial population are eliminated, including:

[0228] Obtain the battery cell supply ratio of each battery cell gear within the production scheduling cycle.

[0229] According to the order scheduling volume of each order and the battery cell gear ratio of each global individual in the first global initial population, the battery cell demand ratio of each global individual in each battery cell gear is obtained respectively.

[0230] According to the battery cell supply ratio and the battery cell demand ratio, a battery cell supply and demand gap corresponding to each global individual is obtained.

[0231] According to the battery supply-demand gap corresponding to each global individual and the battery supply-demand gap range, some global individuals in the first global initial population are eliminated.

[0232] After summarizing all the initial resource allocation strategies in the first stage (such as all Pareto optimal solutions), it is also necessary to consider the supply and demand gap of the battery cells in the overall production scheduling cycle, that is, the solution set obtained in the first stage also needs to be fine-tuned and optimized according to the supply and demand gap of the battery cells. Obtain the battery cell supply ratio of each battery cell gear in the production scheduling cycle. The supply ratio can represent the supply volume. According to the order scheduling volume of each order and the battery cell gear ratio of the global individual, obtain the battery cell demand ratio of the global individual in each battery cell gear. The demand ratio can represent the demand volume. The calculation formula is as follows:

[0233]

[0234]

[0235] Among them, P T is the order scheduling quantity for n orders, and D1×16 is the demand ratio.

[0236] According to the battery supply ratio and battery demand ratio, the battery supply and demand gap corresponding to the global individual can be calculated. The calculation formula is as follows:

[0237] gap=S 1×16 -D 1×16 .

[0238] Among them, gap is the gap between supply and demand of battery cells, S 1×16 The supply ratio.

[0239] According to the battery supply and demand gap of the global individuals and the preset battery supply and demand gap range, the global individuals that do not meet the preset battery supply and demand gap range are eliminated.

[0240] In this embodiment, by excluding infeasible or unreasonable global individuals through preset constraints (i.e., the range of the gap between supply and demand of battery cells), the feasibility of the resource allocation strategy can be ensured, the algorithm convergence can be accelerated, and the quality of the solution can be improved, so as to maximize the balance of supply and demand of photovoltaic modules while maintaining the balance of supply and demand of battery cells.

[0241] In order to make this application easier to understand, the following Figure 9-11 An exemplary application is provided.

[0242] S11: Obtain order information for each order within the production scheduling cycle, wherein the order information may include the required power level of the component, the required quantity, material combination restrictions, battery cell level restrictions, production schedule, benchmark power table, ID list to be predicted, power conversion rules, special requirements, etc.

[0243] S12: Preprocessing data for each order, including: calculating the material combination range and battery cell position range that can be used for each order, and calculating the expected amount of battery cells to be obtained for each battery cell position.

[0244] S13: Multi-objective optimization is performed on each order through genetic algorithm (NSGA-III algorithm) and elite strategy to obtain the Pareto optimal solution (initial resource allocation strategy) of each order.

[0245] The specific process is as follows: encode decision variables (material combination and battery cell gear ratio), set optimization goals (minimize downshift rate and by-product rate), randomly initialize the population based on the material combination range and battery cell gear range, adjust the algorithm's hyperparameters (number of iterations, crossover, mutation probability, etc.), run the genetic algorithm for multi-objective optimization, and obtain the optimization result (Pareto optimal solution).

[0246] S14: Use the Pareto optimal solution of the first stage as the global individual to construct the global initial population, and perform multi-objective optimization under the preset GAP (battery cell supply and demand gap range) through genetic algorithm (NSGA-III algorithm) and elite strategy to obtain the Pareto optimal solution from a global perspective.

[0247] S15: Find the optimal solution (target global individual) according to the preset indicator priority to generate the optimal production strategy (target resource allocation strategy).

[0248] S16: Output the optimal solution and display the obtained optimal solution on the front-end page for instructing / controlling the component production of the factory or workshop.

[0249] In this exemplary application: data preprocessing, business simulation, and multi-objective optimization models are set up, and can adapt to different products and actual business needs, and implement high-efficiency, optimal resource allocation solutions (such as battery cell allocation) for any order; a resource allocation solution that is close to the global optimal solution is used to achieve more refined production resource allocation, less resource mismatch and resource waste, and achieve order delivery with maximized production efficiency and resource conservation; and the supply and demand balance of photovoltaic modules can be met to the greatest extent while maintaining the supply and demand balance of battery cells.

[0250] Embodiment 2

[0251] Fig.12 The block diagram of the resource allocation device for photovoltaic module production according to the second embodiment of the present application is schematically shown. The device can be divided into one or more program modules, one or more program modules are stored in a storage medium, and are executed by one or more processors to complete the embodiment of the present application. The program module referred to in the embodiment of the present application refers to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. Fig.12 As shown, the device 1200 may include: a first acquisition module 1210, a second acquisition module 1220, and an optimization module 1230, wherein:

[0252] The first acquisition module 1210 is used to acquire order information of each order in the production scheduling cycle, wherein the order information includes the required power level of the component;

[0253] The second acquisition module 1220 is used to acquire the available resource range of each order according to the order information;

[0254] The optimization module 1230 is used to perform multi-objective optimization on a plurality of pre-built objective functions based on the available resource range of each order to obtain a target resource allocation strategy, wherein the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle;

[0255] The multiple objective functions include a first objective function for calculating a downshift rate and a second objective function for calculating a byproduct rate, wherein the downshift rate represents a proportion of components higher than the required power gear, and the byproduct rate represents a proportion of components lower than the required power gear.

[0256] As an optional embodiment, the available resource range includes a material combination range and a battery cell gear range.

[0257] As an optional embodiment, the order information further includes material combination restrictions and battery cell position restrictions;

[0258] Correspondingly, the second acquisition module 1220 is further used for:

[0259] Obtaining the supply of various materials within the production scheduling cycle and the supply of battery cells at each battery cell level;

[0260] According to the material combination restrictions of each order, the battery cell position restrictions, the supply of various materials within the production scheduling cycle and the battery cell supply of each battery cell position, the material combination range and battery cell position range of each order are obtained.

[0261] As an optional embodiment, the optimization module 1230 is further configured to:

[0262] Generate an initial resource allocation strategy for each order based on the available resource range of each order;

[0263] Based on the initial resource allocation strategy and preset constraints of each order, the target resource allocation strategy is obtained.

[0264] As an optional embodiment, the optimization module 1230 is further configured to:

[0265] Within the available resource range of the order, a first genetic algorithm is used to generate and iterate N initial populations, each of which includes multiple individuals, each of which includes a material combination and a group of battery cell gear ratios, where N is an integer greater than 1;

[0266] Performing genetic processing on the Nth initial population to obtain a target offspring population corresponding to the Nth initial population;

[0267] Obtaining a target individual in the target offspring population, wherein the target individual is used to generate an initial resource allocation strategy for an order;

[0268] The first genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or a mutation operation.

[0269] As an optional embodiment, the optimization module 1230 is further configured to:

[0270] Generate the first initial population within the available resources of the order;

[0271] Obtaining the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions;

[0272] The second initial population and subsequent initial populations are iterated in sequence: each initial population is obtained by genetically processing the previous initial population based on the downshift rate and byproduct rate of each individual in the previous initial population.

[0273] As an optional embodiment, the optimization module 1230 is further configured to:

[0274] In the process of obtaining one of the target initial populations, a crossover and / or mutation operation is performed on a plurality of individuals in an initial population previous to the target initial population, so as to obtain a descendant population corresponding to the initial population previous to the target initial population;

[0275] Merging the previous initial population of the target initial population and the corresponding offspring population to obtain a merged population;

[0276] Obtaining the downshift rate and byproduct rate corresponding to each individual in the merged population through the multiple objective functions;

[0277] According to the downshift rate and by-product rate corresponding to each individual, some individuals are selected from the merged population to obtain the target initial population.

[0278] As an optional embodiment, the optimization module 1230 is further configured to:

[0279] According to the downshift rate and the byproduct rate corresponding to each individual, the merged population is divided into a plurality of non-dominated layers, each non-dominated layer corresponds to a non-dominated sequence value;

[0280] According to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence.

[0281] As an optional embodiment, the optimization module 1230 is further configured to:

[0282] According to the non-dominated sequence values ​​of the multiple non-dominated layers, some individuals of each non-dominated layer are obtained in sequence to obtain multiple individuals;

[0283] When the number of the plurality of individuals does not satisfy the preset population size, one or more individuals having the minimum vertical distance to the preset reference point are continuously obtained from the non-dominated layer with the minimum non-dominated sequence value.

[0284] As an optional embodiment, the optimization module 1230 is further configured to:

[0285] According to the non-dominated sequence values ​​of the multiple non-dominated layers, multiple individuals satisfying the preset population size are preferentially obtained from the multiple non-dominated layers with small non-dominated sequence values;

[0286] In the last non-dominated layer providing the plurality of individuals, an expected number of individuals are obtained according to the individual crowding distance.

[0287] As an optional embodiment, the order information further includes an order demand and a benchmark power table, the order demand includes the demand of the order at each power level of the component, the each power level includes the required power level, and the benchmark power table includes different benchmark powers of the component under different material combinations, wherein one material combination corresponds to one benchmark power;

[0288] Correspondingly, the optimization module 1230 is also used for:

[0289] According to the individual material combination and the reference power table, obtaining the corresponding reference power;

[0290] Obtaining a component power distribution matrix according to the corresponding reference power and the order demand;

[0291] Obtain the demand-output gap of the component based on the individual material combination, the battery cell gear ratio and the component power distribution matrix;

[0292] According to the demand-output gap, the downshift rate and by-product rate corresponding to the individual are obtained.

[0293] As an optional embodiment, the optimization module 1230 is further configured to:

[0294] The first global initial population is constructed as a global individual using the initial resource allocation strategy of each order, and the second global initial population is iterated sequentially to the Mth global initial population under the preset constraints using the second genetic algorithm; wherein M is an integer greater than 2;

[0295] Based on the preset constraint conditions, performing genetic processing on the Mth global initial population to obtain a target global offspring population corresponding to the Mth global initial population;

[0296] Acquire a target global individual in the target global offspring population, wherein the target global individual is used to generate the target resource allocation strategy;

[0297] The second genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or mutation operation.

[0298] As an optional embodiment, the optimization module 1230 is further configured to:

[0299] Take the initial resource allocation strategy of each order as the global individual and construct the first global initial population;

[0300] According to the preset constraint condition, some global individuals in the first global initial population are eliminated;

[0301] Obtaining the downshift rate and byproduct rate of the remaining global individuals in the first global initial population through the multiple objective functions;

[0302] The second global initial population and the subsequent global initial populations are iterated in sequence: each global initial population is obtained by genetic processing based on the downshift rate and byproduct rate of the remaining global individuals in the previous global initial population.

[0303] As an optional embodiment, the order information further includes order scheduling quantity, and the preset constraint condition includes the gap range between supply and demand of battery cells;

[0304] Correspondingly, the optimization module 1230 is also used for:

[0305] Obtaining the battery cell supply ratio of each battery cell gear within the production scheduling cycle;

[0306] According to the order scheduling volume of each order and the battery cell gear ratio of each global individual in the first global initial population, respectively obtain the battery cell demand ratio of each global individual in each battery cell gear;

[0307] According to the battery cell supply ratio and the battery cell demand ratio, obtaining the battery cell supply and demand gap corresponding to each global individual;

[0308] According to the battery supply-demand gap corresponding to each global individual and the battery supply-demand gap range, some global individuals in the first global initial population are eliminated.

[0309] Embodiment 3

[0310] Fig.13The hardware architecture diagram of a computer device 10000 suitable for implementing a resource configuration method for photovoltaic module production according to the third embodiment of the present application is schematically shown. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle terminal, a virtual device, a workbench, a robot, etc. In other embodiments, the computer device 10000 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server, or a server cluster composed of multiple servers), etc. Fig.13 As shown, the computer device 10000 includes but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can communicate with each other through a system bus. Among them:

[0311] The memory 10010 includes at least one type of computer-readable storage medium, and the readable storage medium includes a flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10010 may be an internal storage module of the computer device 10000, such as a hard disk or a memory of the computer device 10000. In other embodiments, the memory 10010 may also be an external storage device of the computer device 10000, such as a plug-in hard disk equipped on the computer device 10000, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the memory 10010 may also include both the internal storage module of the computer device 10000 and its external storage device. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed on the computer device 10000, such as program code of a resource configuration method for photovoltaic module production, etc. In addition, the memory 10010 can also be used to temporarily store various data that have been output or will be output.

[0312] In some embodiments, the processor 10020 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other chips. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.

[0313] The network interface 10030 may include a wireless network interface or a wired network interface, and the network interface 10030 is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and to establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network may be a wireless or wired network such as an intranet, the Internet, the Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G network, Bluetooth, Wi-Fi, etc.

[0314] It should be pointed out that Fig.13 Only a computer device having components 10010 - 10030 is shown, but it should be understood that implementation of all of the components shown is not a requirement, and more or fewer components may alternatively be implemented.

[0315] In this embodiment, the resource configuration method for photovoltaic module production stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as processor 10020) to complete the embodiment of the present application.

[0316] Embodiment 4

[0317] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the resource configuration method for photovoltaic module production in the embodiment are implemented.

[0318] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the computer-readable storage medium can be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium can also be an external storage device of a computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the computer-readable storage medium can also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the computer-readable storage medium is generally used to store an operating system and various application software installed on the computer device, such as the program code of the resource configuration method for photovoltaic module production in the embodiment. In addition, the computer-readable storage medium can also be used to temporarily store various types of data that have been output or are to be output.

[0319] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present application can be implemented by general-purpose computer devices, they can be concentrated on a single computer device, or distributed on a network composed of multiple computer devices, optionally, they can be implemented by executable program codes of computer devices, so that they can be stored in a storage device and executed by the computer device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0320] It should be noted that the above are only preferred embodiments of the present application, and the patent protection scope of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A resource allocation method for photovoltaic module production, characterized in that: The method comprises: Obtaining order information of each order within a production scheduling cycle, wherein the order information includes a required power level of a component; According to the order information, obtain the available resource range of each order; Generate the first initial population within the available resources of each order; Obtaining the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions; Iterate the second initial population and the subsequent initial populations in sequence: each initial population is obtained by genetically processing the previous initial population based on the downshift rate and byproduct rate of each individual in the previous initial population; each initial population includes multiple individuals, each individual includes a material combination and a group of battery cell gear ratios, where N is an integer greater than 1; Performing genetic processing on the Nth initial population to obtain a target offspring population corresponding to the Nth initial population; Obtaining a target individual in the target offspring population, wherein the target individual is used to generate an initial resource allocation strategy for an order; Wherein, the first genetic algorithm is used to instruct to perform a corresponding genetic process, and the genetic process includes a crossover and / or a mutation operation; Based on the initial resource allocation strategy and preset constraints of each order, the target resource allocation strategy is obtained, where the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle; The multiple objective functions include a first objective function for calculating a downshift rate and a second objective function for calculating a byproduct rate, wherein the downshift rate represents a proportion of components higher than the required power gear, and the byproduct rate represents a proportion of components lower than the required power gear.

2. The method according to claim 1, characterized in that The available resource range includes a material combination range and a battery cell gear range.

3. The method according to claim 2, characterized in that The order information also includes material combination restrictions and battery cell position restrictions; Correspondingly, obtain the available resource range for each order, including: Obtaining the supply of various materials within the production scheduling cycle and the supply of battery cells at each battery cell level; According to the material combination restrictions of each order, the battery cell position restrictions, the supply of various materials within the production scheduling cycle and the battery cell supply of each battery cell position, the material combination range and battery cell position range of each order are obtained.

4. The method according to claim 1, characterized in that: The second initial population and subsequent initial populations are iterated in sequence, including: In the process of obtaining one of the target initial populations, a crossover and / or mutation operation is performed on a plurality of individuals in an initial population previous to the target initial population, so as to obtain a descendant population corresponding to the initial population previous to the target initial population; Merging the previous initial population of the target initial population and the corresponding offspring population to obtain a merged population; Obtaining the downshift rate and byproduct rate corresponding to each individual in the merged population through the multiple objective functions; According to the downshift rate and by-product rate corresponding to each individual, some individuals are selected from the merged population to obtain the target initial population.

5. The method according to claim 4, characterized in that Selecting some individuals from the merged population according to the downshift rate and byproduct rate corresponding to each individual, including: According to the downshift rate and the byproduct rate corresponding to each individual, the merged population is divided into a plurality of non-dominated layers, each non-dominated layer corresponds to a non-dominated sequence value; According to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence.

6. The method according to claim 5, characterized in that According to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence, including: According to the non-dominated sequence values ​​of the multiple non-dominated layers, some individuals of each non-dominated layer are obtained in sequence to obtain multiple individuals; When the number of the plurality of individuals does not satisfy the preset population size, one or more individuals having the minimum vertical distance to the preset reference point are continuously obtained from the non-dominated layer with the minimum non-dominated sequence value.

7. The method according to claim 5, characterized in that According to the non-dominated sequence value, individuals for constructing the target initial population are obtained from a plurality of non-dominated layers in sequence, including: According to the non-dominated sequence values ​​of the multiple non-dominated layers, multiple individuals satisfying the preset population size are preferentially obtained from the multiple non-dominated layers with small non-dominated sequence values; In the last non-dominated layer providing the plurality of individuals, an expected number of individuals are obtained according to the individual crowding distance.

8. The method according to claim 1, characterized in that The order information also includes an order demand and a benchmark power table, wherein the order demand includes the demand for each power level of the component, each power level includes the required power level, and the benchmark power table includes different benchmark powers of the component under different material combinations, wherein one material combination corresponds to one benchmark power; Correspondingly, the downshift rate and byproduct rate of each individual in the first initial population are obtained, including: According to the individual material combination and the reference power table, obtaining the corresponding reference power; Obtaining a component power distribution matrix according to the corresponding reference power and the order demand; Obtain the demand-output gap of the component based on the individual material combination, the battery cell gear ratio and the component power distribution matrix; According to the demand-output gap, the downshift rate and by-product rate corresponding to the individual are obtained.

9. The method according to any one of claims 1 to 8, characterized in that: Based on the initial resource allocation strategy and preset constraints of each order, the target resource allocation strategy is obtained, including: The first global initial population is constructed as a global individual using the initial resource allocation strategy of each order, and the second global initial population is iterated sequentially to the Mth global initial population under the preset constraints using the second genetic algorithm; wherein M is an integer greater than 2; Based on the preset constraint conditions, performing genetic processing on the Mth global initial population to obtain a target global offspring population corresponding to the Mth global initial population; Acquire a target global individual in the target global offspring population, wherein the target global individual is used to generate the target resource allocation strategy; The second genetic algorithm is used to instruct execution of a corresponding genetic process, wherein the genetic process includes a crossover and / or mutation operation.

10. The method according to claim 9, characterized in that The first global initial population is constructed as a global individual using the initial resource allocation strategy of each order, and the second global initial population is iterated sequentially to the Mth global initial population under the preset constraints using the second genetic algorithm, including: Take the initial resource allocation strategy of each order as the global individual and construct the first global initial population; According to the preset constraint condition, some global individuals in the first global initial population are eliminated; Obtaining the downshift rate and byproduct rate of the remaining global individuals in the first global initial population through the multiple objective functions; The second global initial population and the subsequent global initial populations are iterated in sequence: each global initial population is obtained by genetic processing based on the downshift rate and byproduct rate of the remaining global individuals in the previous global initial population.

11. The method according to claim 10, characterized in that The order information also includes the order scheduling quantity, and the preset constraint condition includes the gap range between supply and demand of battery cells; Correspondingly, according to the preset constraint condition, some global individuals in the first global initial population are eliminated, including: Obtaining the battery cell supply ratio of each battery cell gear within the production scheduling cycle; According to the order scheduling volume of each order and the battery cell gear ratio of each global individual in the first global initial population, respectively obtain the battery cell demand ratio of each global individual in each battery cell gear; According to the battery cell supply ratio and the battery cell demand ratio, obtaining the battery cell supply and demand gap corresponding to each global individual; According to the battery supply-demand gap corresponding to each global individual and the battery supply-demand gap range, some global individuals in the first global initial population are eliminated.

12. A resource allocation device for photovoltaic module production, characterized in that: The device comprises: A first acquisition module is used to acquire order information of each order in a production scheduling cycle, wherein the order information includes a required power level of a component; A second acquisition module, used to acquire the available resource range of each order according to the order information; An optimization module is used to generate a first initial population within the available resource range of each order; obtain the downshift rate and byproduct rate of each individual in the first initial population through the multiple objective functions; iterate the second initial population and the subsequent initial populations in sequence: each initial population is obtained by genetically processing the previous initial population based on the downshift rate and byproduct rate of each individual in the previous initial population; each initial population includes multiple individuals, each individual includes a material combination and a group of battery cell gear ratios, wherein N is an integer greater than 1; perform genetic processing on the Nth initial population to obtain a target offspring population corresponding to the Nth initial population; obtain a target individual in the target offspring population, and the target individual is used to generate an initial resource allocation strategy for an order; wherein the first genetic algorithm is used to instruct the execution of corresponding genetic processing, and the genetic processing includes crossover and / or mutation operations; based on the initial resource allocation strategy of each order and preset constraints, obtain the target resource allocation strategy; Among them, the target resource allocation strategy is used to indicate the resource allocation within the production scheduling cycle, and the multiple objective functions include a first objective function for calculating the downshift rate and a second objective function for calculating the by-product rate. The downshift rate represents the proportion of components higher than the required power level, and the by-product rate represents the proportion of components lower than the required power level.

13. A computer device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein: The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method according to any one of claims 1 to 11 is implemented.

Images