Picking and sorting decision-making method considering sowing wall capacity, service platform and storage medium

Abstract

The invention relates to a selection and sorting decision-making method considering the sowing wall capacity, a service platform and a storage medium, and the method comprises the steps: carrying out the random coding and population initialization of SKUs associated with a plurality of obtained to-be-selected selection orders according to a preset coding mode, and generating a plurality of first sorting coding bodies; using a heap optimization simulated annealing algorithm to perform sorting correction preprocessing on the sub-codes corresponding to the plurality of first sorting coding bodies to generate a plurality of second sorting coding bodies, and based on the plurality of second sorting coding bodies, repeatedly performing non-dominated solution sorting and genetic evolution operation iteration by using an NSGA-II algorithm and the fitness of the corresponding sorting coding bodies, so as to obtain a sorting result; until a plurality of alternative sorting coding bodies are generated; and selecting a target sorting coding body from the plurality of alternative sorting coding bodies according to the fitness of the alternative sorting coding bodies, and taking a plurality of target sorting batches obtained by processing the target sorting coding body according to a sorting decision rule and sorting paths mapped by the target sorting batches as decision results.

Description

Technical Field

[0001] The present application relates to the field of computer intelligent application technology, and in particular to a picking and sorting decision method, service platform, and storage medium that considers the capacity of a seed wall. Background Art

[0002] In the existing technology, since e-commerce warehouses usually face large order volumes, high timeliness requirements, and high picking density, in order to improve efficiency, most of them adopt a multi-person collaborative picking operation mode. However, when batch picking operations are carried out simultaneously and sorting or packaging resources are limited, if the batch workload is unbalanced, it may cause coordination problems between order operations and generate waiting time; in related technologies, research on the order sorting link usually assumes that there is no capacity limit of the sorting area when executing the operation, or assumes that the capacity of the sorting area is large enough. However, in actual warehouse operations, equipment such as seed walls used in order sorting operations are capacity-limited. If the picking plan does not take into account the resource limitations of the sorting operation link, congestion and waiting will occur in the sorting operation.

[0003] Currently, the picking planning in related technologies does not take into account the resource limitations of the sorting operation, and congestion and waiting problems are prone to occur during the sorting operation. No effective solution has been proposed yet. Summary of the Invention

[0004] The embodiments of the present application provide a picking and sorting decision method, service platform and storage medium that take into account the capacity of the seed wall, so as to at least solve the problem in the related art that the picking planning does not take into account the resource limitations of the sorting operation link, and congestion and waiting are prone to occur during the sorting operation.

[0005] In the first aspect, an embodiment of the present application provides a picking and sorting decision method that takes into account the capacity of the seed wall, comprising: after obtaining a plurality of picking orders to be picked that are trapped in a picking order pool corresponding to a target warehouse, the SKUs associated with the plurality of picking orders are randomly encoded and the population is initialized according to a preset encoding method to generate a plurality of first sorting coding bodies, wherein each of the picking orders is associated with at least one SKU, and the first sorting coding body includes a plurality of subcodes that represent the picking order with gene bits, and each subcode corresponds to one SKU; using a heap optimized simulated annealing algorithm, the subcodes corresponding to the plurality of the first sorting coding bodies are sorted and corrected preprocessed to generate a plurality of second sorting coding bodies, and based on the plurality of the second sorting coding bodies, the NSGA-Ⅱ algorithm and the corresponding sorting coding are repeatedly executed. The fitness of the code body is determined by performing non-dominated solution sorting and genetic evolution operations iteratively until multiple alternative sorting code bodies are generated, wherein the fitness is determined according to the objective function values ​​corresponding to the multiple picking batches corresponding to the corresponding sorting code body, and the picking batches are generated by picking and sorting the SKUs of the corresponding sorting code body according to the picking and sorting decision rules considering the seed wall capacity, and the genetic evolution operation adopts adaptive crossover probability and adaptive mutation probability; according to the fitness of the alternative sorting code body, a target sorting code body is selected from the multiple alternative sorting code bodies, and the multiple target picking batches obtained by processing the target sorting code body according to the picking and sorting decision rules, and the picking path mapped by the sorting of all the SKUs of each target picking batch is used as the decision result.

[0006] In a second aspect, an embodiment of the present application provides a service platform, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps of the picking and sorting decision method considering the seed wall capacity described in the first aspect above.

[0007] In a third aspect, an embodiment of the present application provides a storage medium on which a computer program is stored. When the program is executed by a processor, the picking and sorting decision method considering the seed wall capacity as described in the first aspect above is implemented.

[0008] Compared with the related art, the picking and sorting decision-making method, service platform and storage medium provided by the embodiments of the present application consider the capacity of the sowing wall. After obtaining multiple picking orders to be picked in the picking order pool corresponding to the target warehouse, according to a preset coding method, the SKUs associated with the multiple picking orders are randomly coded and the population is initialized to generate multiple first sorting coding bodies; using the heap optimization simulated annealing algorithm, the sub-codes corresponding to the multiple first sorting coding bodies are sorted and preprocessed to generate multiple second sorting coding bodies, and based on the multiple second sorting coding bodies, the non-dominated sorting and genetic evolution operation iterations are repeatedly performed using the NSGA-II algorithm and the fitness of the corresponding sorting coding body until multiple alternative sorting coding bodies are generated. The fitness is determined according to the objective function values corresponding to the multiple picking batches corresponding to the corresponding sorting coding body. The picking batches are generated by picking and sorting the SKUs of the corresponding sorting coding body according to the picking and sorting decision rule considering the capacity of the sowing wall; according to the fitness of the alternative sorting coding body, a target sorting coding body is selected from the multiple alternative sorting coding bodies, and the multiple target picking batches obtained by processing the target sorting coding body according to the picking and sorting decision rule, and the picking path mapped by the sorting of all the SKUs of each target picking batch are used as the decision result. By introducing the capacity constraint of the sowing wall and using preset multi-objectives as the fitness, the joint optimization decision of picking and sorting considering the order picking path, the capacity of the sowing wall and the capacity constraint of the picking equipment is carried out. Under the limited capacity of the sowing wall, the joint optimization of picking and sorting operations is realized, the backlog and waiting of orders in the sorting link are avoided, the efficient progress of order operations is ensured, the overall operation efficiency is improved, and the problem that the picking plan in the related art does not consider the resource limitation in the sorting operation link and congestion and waiting are prone to occur in the sorting operation is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application, and do not constitute an improper limitation to the present application. In the drawings: Figure 1 is the hardware structure block diagram of the terminal of the picking and sorting decision-making method considering the capacity of the sowing wall according to the embodiment of the present application; Figure 2 is the flowchart of the picking and sorting decision-making method considering the capacity of the sowing wall according to the embodiment of the present application; Figure 3 is the structure block diagram of the picking and sorting decision-making device considering the capacity of the sowing wall according to the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and examples. 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 provided in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for ordinary technicians in the field related to the contents disclosed in the present application, some changes such as design, manufacturing or production based on the technical contents disclosed in the present application are only conventional technical means and should not be understood as the contents disclosed in the present application being insufficient.

[0011] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it refer to independent or alternative embodiments that are mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments unless there is a conflict.

[0012] Unless otherwise defined, technical or scientific terms used in this application shall have the ordinary meaning as understood by persons of ordinary skill in the art to which this application belongs. The use of "a," "an," "an," "the," and similar expressions in this application does not denote a limitation of quantity and may refer to either the singular or the plural. The terms "comprise," "include," "have," and any variations thereof, as used in this application, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device comprising a series of steps or modules (units) is not limited to the listed steps or units but may also include steps or units not listed, or may include other steps or units inherent to the process, method, product, or device. As used in this application, "multiple steps" means two or more steps. "And / or" describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" may mean: A exists alone, A and B exist simultaneously, or B exists alone. The terms "first," "second," and "third," etc., as used in this application, simply distinguish similar objects and do not imply a specific ordering of the objects.

[0013] Before describing the method of the embodiment of the present application, the related technologies involved in the embodiment of the present application are described as follows:

[0014] The Non-dominated Sorting Genetic Algorithm II (NAGA-II) algorithm is a multi-objective optimization algorithm based on the genetic algorithm (GA). It is an improved version of NAGA (Non-dominated Sorting Genetic Algorithm) specifically designed for non-dominated sorting problems. NSGA-II inherits the principles of the genetic algorithm (GA), retaining operations such as selection, crossover, and mutation. However, it incorporates mechanisms such as non-dominated sorting and crowding comparison specifically for multi-objective optimization problems. By simulating the natural process of biological evolution, NSGA-II is able to find uniformly distributed, non-dominated Pareto optimal solutions in multi-objective optimization. The principle of the NSGA-II algorithm is to optimize multi-objective problems by simulating the natural processes of biological reproduction, survival of the fittest, and elimination of the weak. In the NSGA-II algorithm, each individual represents a spatial solution, just like an individual organism in nature, with its own unique characteristics (encoded by chromosome genes). These individuals are distributed in the search space and continuously optimized through evolutionary operations. The process of finding the multi-objective Pareto optimal solution set in the iterative process is analogous to the process of survival of the fittest among individuals in a population.

[0015] The specific process of the NSGA-Ⅱ algorithm is as follows: 1. Initialize the population, randomly generate an initial population of size M, initialize the maximum number of iterations, crossover rate and mutation rate, etc.; 2. Fast non-dominated sorting, calculate the fitness function of the initialized population and sort it; 3. Select, crossover, and mutate to obtain the first-generation offspring population and enrich the diversity of the population; 4. The parent and offspring populations are merged, and the population size changes from M to 2M; 5. Fast non-dominated sorting, perform fast non-dominated sorting on the new population; 6. Calculate the crowding degree; 7. Form a new parent generation, select suitable individuals to form the new parent generation based on the non-dominated relationship and the crowding degree of the individuals; 8. Select, crossover, and mutate, and repeat the cycle until the iteration criteria are met.

[0016] NSGA-II mainly uses fast non-dominated sorting and crowding calculation to screen individuals, thereby ensuring the quality of the Pareto optimal solution and maintaining the diversity of the population during the optimization process. Fast Non-dominated Sorting (FSS) is one of the core components of the NSGA-II algorithm. It assigns each individual in a population to a different level (Front) based on the Pareto dominance relationship between populations. This ensures that the best solutions are retained first and the diversity of the population is maintained during the multi-objective optimization process.

[0017] Non-dominated sorting defines the dominance count of each individual in a population. In multi-objective optimization, for individual p to dominate individual q, the following two conditions must be met: p is no worse than q on all objectives; and p is better than q on at least one objective. For example, in a three-objective minimization problem, if the three objective values ​​corresponding to individual p are [4, 1, 6] and the three objective values ​​corresponding to individual q are [2, 1, 5], then all objectives of individual q are less than or equal to the objective values ​​of individual p, and the first and third objectives of individual q are less than the objective values ​​of individual p, satisfying the dominance relationship, and therefore individual q dominates individual p. By traversing the dominance relationships between individuals in the population, the entire population is stratified according to the dominance relationships. The first layer (Front 1) contains solutions not dominated by any individual, which is called the optimal solution set; the second layer (Front 2) contains solutions dominated only by individuals in the first layer; the third layer (Front 3) contains solutions dominated only by individuals in the first two layers, and so on. Ultimately, the entire population is divided into multiple non-dominated layers, each corresponding to a set of solutions with the same dominance rank.

[0018] The crowding distance calculation measures the sparseness of each individual in the target space, allowing for comparison of non-dominated solutions. This allows for prioritizing more dispersed solutions when selecting individuals for the next generation, preventing the population from converging to a narrow region. Generally speaking, populations with large crowding distances are preferred because their solutions are more dispersed, making it easier to obtain a complete Pareto frontier and thus maintaining population diversity.

[0019] The List-Based Simulated Annealing (LBSA) algorithm is an improved version of the original simulated annealing (SA) algorithm. The main difference between the LBSA algorithm and the SA algorithm lies in the temperature control strategy. The LBSA algorithm determines a state transition probability P0 at the beginning and infers a temperature inversion formula based on the corresponding acceptance probability formula of the SA algorithm: , an initial temperature list is first generated by the temperature inversion formula, and in the subsequent simulated annealing, the preferred annealing is performed according to the temperature stored in this list. The specific process of LBSA is as follows: step (1) initialize the temperature list and the number of iterations; step (2) take the maximum value tmax in the list, update the number of outer loop iterations, temperature, number of inner loop iterations, and the number of accepted difference solutions; step (3) create a new solution; (4) judge and compare, if the new solution is better, go to step (6); if the new solution is worse, go to step (5); step (5) calculate the acceptance probability, judge whether to accept, if accepted, update t and c, go to step (6); if not accepted, go to step (7); step (6), replace the original solution; step (7) judge whether it is less than the number of inner loop iterations, if so, go to step (3); if not, go to step (8); step (8) judge whether the number of accepted difference solutions is 0, if not, update the temperature list; step (9) judge whether it is less than the number of outer loop iterations, if so, go to step (2); if not, end.

[0020] A Stock Keeping Unit (SKU) is the smallest available unit for inventory control. It's typically measured in units like pieces, boxes, and pallets, and is used to number and categorize products in logistics management. Each product has a unique SKU number, which includes detailed product attributes such as brand, model, configuration, packaging capacity, unit, production date, expiration date, intended use, price, and origin. SKUs also play a crucial role in warehouse management. By numbering and managing SKUs, warehouse layout and storage strategies can be optimized, reducing inefficient movement and routing conflicts, and improving operational efficiency.

[0021] The following describes a specific embodiment of the picking and sorting decision method considering the sowing wall capacity according to an embodiment of the present application. The method embodiment provided in this embodiment can be executed in a terminal, a computer or a similar computing device. Taking running on a terminal as an example, Figure 1 This is a hardware structure block diagram of the terminal of the picking and sorting decision method considering the seed wall capacity in the embodiment of the present application. Figure 1 As shown, the terminal may include one or more ( Figure 1 Only one is shown) a processor 102 (the processor 102 may include but is not limited to a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above terminal. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0022] Memory 104 can be used to store computer programs, such as application software programs and modules, such as the computer program corresponding to the picking and sorting decision method considering seed wall capacity in the embodiments of the present invention. Processor 102 executes the computer programs stored in memory 104 to execute various functional applications and data processing, thereby implementing the aforementioned methods. Memory 104 can include high-speed random access memory (RAM) and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 104 may further include memory located remotely from processor 102, which can be connected to terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0023] Transmission device 106 is used to receive or transmit data via a network. A specific example of the aforementioned network may include a wireless network provided by the communications provider of terminal 10. In one embodiment, transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0024] This embodiment provides a picking and sorting decision method that considers the seed wall capacity and runs on the above terminal. Figure 2 FIG. 1 is a flowchart of a picking and sorting decision method considering the sowing wall capacity according to an embodiment of the present application. Figure 2 As shown, the process includes the following steps:

[0025] Step S201, after obtaining multiple picking orders to be picked that are trapped in the picking order pool corresponding to the target warehouse, randomly encode and initialize the SKUs associated with the multiple picking orders according to a preset encoding method to generate multiple first sorting code bodies, wherein each picking order is associated with at least one SKU, and the first sorting code body includes multiple sub-codes that represent the picking order with gene bits, and each sub-code corresponds to one SKU.

[0026] In this embodiment, the execution subject of the embodiment of the present application includes but is not limited to the warehouse order management system. After receiving the customer order, the execution subject will convert the received order into a picking order and intercept it in the picking order pool to wait for picking planning; after starting the picking planning, it will first extract multiple picking orders currently to be planned from the picking order pool, and obtain all the SKUs to be coded based on the SKUs contained in each picking order. For example: there are picking orders a[1,2,3,4] and picking orders b[5,6,7] and picking orders c[8,9], then the corresponding SKUs to be coded are: 1 ,2,3,4,5,6,7,8,9; At this time, the SKU to be encoded is encoded according to the preset encoding method. It can be understood that encoding is the process of mapping the feasible solution of the problem from the solution space to the search space that the algorithm can handle. Common encoding methods include natural number encoding and binary encoding. In this embodiment, the batching and picking order of items in the order (mapped to the corresponding SKU) are decided, and each SKU has a unique natural number number. The final solution is a string of natural number sequences with a certain order. In this embodiment, natural number encoding is used to improve the efficiency of the algorithm and the quality of the solution.

[0027] In this embodiment, the SKUs associated with multiple picking orders are encoded to generate multiple first sorting code bodies, that is, the corresponding initial populations. For example, the orders to be picked include: picking order 1 (with SKU numbers: 1, 2, 3), picking order 2 (with SKU numbers: 4, 5, 6), picking order 3 (with SKU numbers: 7, 8), picking order 4 (with SKU numbers: 9, 10, 11, 12), picking order 5 (with SKU numbers: U numbers are: 13, 14), picking order 6 (with SKU numbers: 15, 16), picking order 7 (with SKU numbers: 17, 18, 19, 20), picking order 8 (with SKU numbers: 21, 22, 23), picking order 9 (with SKU numbers: 24, 25, 26), picking order 10 (with SKU numbers: 27, 28), after encoding according to the natural number encoding method, the first sorting code body 1 is generated: [10, 1, 11, 27, 28, 6, 4, 9, 14, 22, 15, 25, 16, 3, 20, 24, 19, 26, 18, 23, 17, 5, 13, 21, 8, 7, 12, 2], and then randomly initialized. The generated partial sorting code body includes: the first sorting code body 2 [13, 1, 8, 27, 28, 6, 4, 9, 14, 3, 15, 25, 16, 22, 20, 24, 26, 18, 23, 17, 19, 5, 10, 21, 11, 7, 12, 2], the first sorting code body 3 [27, 28, 6, 10, 1, 11, 4, 9, 14, 22, 15, 24, 19, 26, 18, 23, 25, 16, 3, 20, 17, 5, 13, 21, 8, 7, 12, 2]; It can be understood that in a first sorting code body, the position of a SKU indicates the order in which it is picked. For example, 10 in the first sorting code body is in the first order. At this time, it means that the SKU will be picked by the first index, but it does not mean that it will be assigned to the corresponding picking batch.

[0028] Step S202: Using the heap optimized simulated annealing algorithm, the subcodes corresponding to the multiple first sorting code bodies are sorted and corrected preprocessed to generate multiple second sorting code bodies. Based on the multiple second sorting code bodies, the NSGA-Ⅱ algorithm and the fitness of the corresponding sorting code bodies are repeatedly executed to perform non-dominated solution sorting and genetic evolution operations until multiple alternative sorting code bodies are generated. The fitness is determined according to the objective function values ​​corresponding to the multiple picking batches corresponding to the corresponding sorting code bodies. The picking batches are generated by picking and sorting the SKUs of the corresponding sorting code bodies according to the picking and sorting decision rules considering the seed wall capacity. The genetic evolution operation adopts adaptive crossover probability and adaptive mutation probability.

[0029] In this embodiment, due to the initialization of the first sorting code body, the corresponding solutions are unevenly distributed, the initial population quality is poor, which affects the subsequent NSGA-Ⅱ algorithm for picking and sorting joint optimization effect and convergence efficiency, and the solution is prone to fall into the local optimum. In this embodiment, the LBSA algorithm (heap optimization simulated annealing algorithm) is used to pre-process and optimize the initial population, which can effectively improve the quality of the initial population and avoid premature convergence to the local optimum. In this way, the NSGA-Ⅱ algorithm can start optimization from a better solution in the subsequent evolution process, accelerate the expansion of the Pareto frontier, and enhance the diversity of solutions.

[0030] In this embodiment, an algorithm combining NSGA-II and LBSA is adopted, and the specific process includes the following steps:

[0031] Step 1: Initialize parameters, including the size of the population composed of the first sorted encoding body, the maximum number of iterations, the adaptive crossover probability and the adaptive mutation probability, the list length of LBSA, and the number of internal and external loops.

[0032] Step 2: Initialize the population and randomly generate an initial population of size M, including multiple first-ordered encoding bodies.

[0033] Step 3: Use the LBSA algorithm to preprocess the initial population to obtain the second sorting code body, and then use fast non-dominated sorting to calculate the fitness of the optimized initialization population and sort it.

[0034] Step 4: After the selection is completed, crossover and mutation are performed with the preset adaptive crossover probability and adaptive mutation probability to obtain the first generation offspring population.

[0035] Step 5: The parent population (initial population) and the offspring population (first generation offspring population) are merged, and the population size changes from M to 2M. The merged population is quickly non-dominated sorted to obtain a new population with the preset population size, and a population composed of the corresponding sorting code bodies is obtained.

[0036] Step 6: Calculate the congestion level.

[0037] Step 7: Form a new parent generation. Select appropriate individuals to form the new parent generation based on the non-dominated relationship and the crowding degree of the individuals.

[0038] Step 8: Repeat the selection, crossover, and mutation cycles until the iteration criteria are met.

[0039] It can be understood that the operation steps corresponding to the LBSA and NSGA-Ⅱ algorithms used in this embodiment are known or clear to those skilled in the art. The relevant operations of the existing LBSA and NSGA-Ⅱ algorithms can be used to achieve preprocessing optimization of the first sorting code body, and to achieve non-dominated solution sorting and congestion calculation based on the second sorting code body, and to perform corresponding genetic evolution operations of selection, crossover and mutation, so as to obtain alternative sorting code bodies that meet the preset standards.

[0040] In this embodiment, in the process of calculating the fitness, the relevant sorting code bodies are first picked and sorted according to the picking and sorting decision rules taking into account the seed wall capacity to generate corresponding picking batches. At the same time, based on the SKU sorting of each picking batch, the corresponding picking path is determined. Then, according to each picking batch and the corresponding picking path and the target values ​​corresponding to the preset multiple objective functions, the picking time corresponding to each picking batch, the sorting completion time of the order and the picking operation balance are determined to determine the fitness of each corresponding sorting code body. Then, according to the fitness, the non-dominated solution sorting and genetic evolution operation iterations are performed until the corresponding fitness guides the genetic evolution and determines the alternative sorting code body that meets the set requirements.

[0041] In step S203, a target sorting code body is selected from multiple candidate sorting code bodies according to the fitness of the candidate sorting code bodies, and multiple target picking batches obtained by processing the target sorting code body according to the picking and sorting decision rules, and the picking path mapped by the sorting of all SKUs in each target picking batch are used as the decision result.

[0042] In this embodiment, the LBSA is combined with the NSGA-II algorithm to perform multiple picking and sorting joint optimizations on the first sorting code body of the code until the fitness of the corresponding sorting code body meets the set requirements or the number of iterations reaches the set requirements. At this time, multiple candidate sorting code bodies are obtained, and then the corresponding candidate sorting code body is selected from the multiple candidate sorting code bodies that meet the requirements to obtain the target sorting code body. Because before calculating the fitness of the corresponding sorting code body, the sorting code body has been sorted and processed according to the picking and sorting decision rules considering the seed wall capacity, that is, the corresponding picking batch is generated. When the corresponding target sorting code body is determined, the picking batch corresponding to the target sorting code body is also determined, that is, the corresponding multiple target picking batches and the picking path mapped by the sorting of the SKUs of each target picking batch are obtained, and the decision results of order batching and picking path planning are obtained.

[0043] Through the above steps S201 to S203, after obtaining multiple picking orders to be picked in the picking order pool corresponding to the target warehouse, the SKUs associated with the multiple picking orders are randomly encoded and the population is initialized according to the preset encoding method to generate multiple first sorting code bodies; the heap optimization simulated annealing algorithm is used to perform sorting correction preprocessing on the subcodes corresponding to the multiple first sorting code bodies to generate multiple second sorting code bodies, and based on the multiple second sorting code bodies, the NSGA-Ⅱ algorithm and the fitness of the corresponding sorting code body are repeatedly executed to perform non-dominated solution sorting and genetic evolution operation iterations until multiple alternative sorting code bodies are generated. The fitness is determined according to the objective function values ​​corresponding to the multiple picking batches corresponding to the corresponding sorting code body. The picking batch is determined according to the picking and sorting decision rule considering the seed wall capacity. SKUs of the selected items are picked and sorted according to the fitness of the selected sorting code body; a target sorting code body is selected from multiple alternative sorting code bodies according to the fitness of the alternative sorting code body, and multiple target picking batches obtained by processing the target sorting code body according to the picking and sorting decision rules, and a picking path mapped by the sorting of all SKUs in each target picking batch are used as the decision result. By introducing the capacity constraint of the seed wall and taking the preset multiple objectives as the fitness, a picking and sorting joint optimization decision is made considering the order picking path, seed wall capacity and picking equipment capacity constraints. Under the limited seed wall capacity, the picking and sorting operations are jointly optimized to avoid the backlog and waiting of orders in the sorting link, ensure the efficient execution of order operations, improve the overall operation efficiency, and solve the problem that the picking planning in the related technology does not take into account the resource limitation of the sorting operation link, and the sorting operation is prone to congestion and waiting.

[0044] In some embodiments, before determining the fitness of the corresponding sorting code body, the following steps are further performed to implement picking and sorting processing of the corresponding sorting code body according to the picking and sorting decision rule considering the sowing wall capacity:

[0045] Step 21, determine the active orders currently associated with the target seed wall and the picking capacity corresponding to each picking batch to be constructed, where the active orders are used to represent the picking orders allowed to be loaded on the target seed wall, and the picking capacity is used to represent the number of SKUs picked by the picking equipment associated with the target seed wall for a picking batch.

[0046] In this embodiment, during the process of picking and sorting the corresponding SKUs, the capacity of the seed wall is taken into consideration, and the priority of the orders to be picked is determined by the capacity of the seed wall. In order to take into account the capacity of the seed wall, when indexing and picking the corresponding SKU, the seed wall will activate or associate at least one active order, and the currently activated or associated order has priority when indexing and picking the corresponding SKU. That is, when the SKU indexed and picked belongs to the SKU of the picking order corresponding to the active order currently associated with the seed wall, the SKU is allowed to be batched into the corresponding picking batch, otherwise, the SKU will be skipped; and the picking capacity of each picking batch is determined according to the capacity corresponding to the picking equipment of the seed wall. For example: if the capacity of the picking equipment associated with the seed wall is 4, then when picking the current picking batch, the upper limit of the number of SKUs picked is 4, and among all SKUs of the corresponding active order, the upper limit of the number of SKUs picked is 4. When all SKUs are picked, if there are less than 4 SKUs in the corresponding picking batch, the SKUs of other picking orders will not be added, and the state of less than 4 SKUs will be maintained to ensure that the picked picking batch is feasible; it can be understood that when executing the picking and sorting of the corresponding sorting code body, it is necessary to first determine the active orders associated with the target seed wall and the picking capacity of the corresponding picking equipment to determine the order of picking the picking orders and the batch parameters of the picking batches.

[0047] Step 22: perform index picking on all SKUs of the third sorting code body currently to be picked and sorted, and determine whether the currently picked SKU belongs to the SKU of the currently associated active order.

[0048] Step 23: When it is determined that the currently picked SKU belongs to the SKU of the currently associated active order, the currently picked SKU is added to an unfilled picking batch.

[0049] In this embodiment, when traversing all SKUs in the third sorting code body, only if the order to which the SKU belongs already exists in the active order list (corresponding to a SKU in the active order), the SKU will be placed in a currently unfilled picking batch; otherwise, the SKU picked by the index will be skipped and the next SKU will be read;

[0050] Step 24, repeatedly execute the steps of index picking corresponding to the SKU, determining whether the picked SKU is the SKU corresponding to the active order, and adding the SKU to the corresponding picking batch, until multiple picking batches corresponding to the currently associated active order are decided, and the multiple picking batches decided by multiple decisions are combined into a total picking batch, wherein the multiple picking batches corresponding to the third sorting code body include the total picking batch, and the number of SKUs added in each picking batch does not exceed the picking capacity.

[0051] In this embodiment, after a SKU is picked, the corresponding SKU capacity will be added to the current load of the picking device associated with the seed wall. When the load of the picking device reaches its maximum capacity, the current picking batch is terminated, and the picking object returns to the seed wall to sort the picked SKUs. After that, a new picking batch will continue to be generated from the position where the previous picking batch ended at the third sorting code body until the index picking of a third sorting code body is completed.

[0052] Through the above steps 21 to 24, by considering the capacity constraints of the seed wall and the capacity constraints of the picking equipment, priority picking is performed on the active orders associated with the seed wall, and the corresponding sorting code bodies are picked and sorted according to the picking and sorting decision rules considering the seed wall capacity, ensuring the feasibility of picking batches and picking path planning, optimizing overall operation efficiency, and improving order operation continuity.

[0053] In some embodiments, when it is determined that the currently picked SKU does not belong to the SKU of the currently associated active order, the following steps are further performed:

[0054] Step 31 : Determine whether the target seed wall allows adding associated active orders based on the number of currently associated active orders and the order loading capacity corresponding to the target seed wall.

[0055] In this embodiment, if it is determined whether the order to which the SKU currently indexed and picked belongs has been activated (that is, it is not an SKU belonging to an active order), then it is first determined whether there is any remaining order capacity in the seed wall, that is, whether there is any remaining order capacity of an active order that can be activated or associated. For example, when the order capacity of the active order associated with the seed wall is 2, and an active order has been associated with it before, then there is an order capacity of an active order that can be associated.

[0056] Step 32: When it is determined that the target seed wall allows adding associated active orders, the picking order to which the currently picked SKU belongs is used as the added active order, and the associated active order and all added active orders are used as the currently associated active orders of the target seed wall, and the currently picked SKU is added to an unfilled picking batch.

[0057] In this embodiment, when it is determined that there is still order capacity for active orders that can be associated, the picking order to which the currently picked SKU belongs will be used as the active order associated with the target seed wall, that is, the picking order to which the currently picked SKU belongs will be activated, and the currently picked SKU will be added to a corresponding picking batch.

[0058] Step 33: When it is determined that the target seed wall does not allow adding associated active orders, index picking is performed on the next SKU.

[0059] In this embodiment, when it is determined that there is no order capacity for an associated active order, the currently picked SKU will be skipped, and the next SKU needs to be indexed and picked to determine whether the next SKU belongs to the SKU of the currently activated active order.

[0060] In some embodiments, after determining the multiple picking batches corresponding to the currently associated active order, the following steps are further performed:

[0061] Step 41 , index and pick the next SKU, and determine the picking order to which the next SKU belongs.

[0062] Step 42: Release all currently associated active orders, associate at least the picking order to which the next SKU belongs as the corresponding active order, and add the next SKU picked by the index to a newly constructed picking batch.

[0063] In this embodiment, after all SKUs of the active orders associated with the target seed wall are indexed and picked out, the next SKU is indexed and picked. This SKU must not belong to the picking batch that can be batched. At this time, the associated active order is released, that is, the order capacity of the target seed wall is released, and the picking order of the next SKU picked by the associated index is activated, that is, at least one active order is activated, and then a new picking batch is constructed, and the next SKU is added to the picking batch. It can be understood that when continuing to index and pick candidate SKUs, the target seed wall may not activate or associate the corresponding active order before indexing a SKU that does not belong to at least one activated active order. When a SKU that does not belong to at least one activated active order is indexed, the picking order to which the SKU picked by activating the index belongs is the active order, so as to ensure the feasibility of the picking batch sorted out.

[0064] In some embodiments, after the SKUs picked for the current index are added to the corresponding picking batch and before the multiple picking batches determined by multiple decisions are combined into a total picking batch, the following steps are further performed:

[0065] Step 51: determine whether the currently selected SKU is at the end of the third sorting code body.

[0066] Step 52: When it is determined that the currently picked SKU is at the end of the third sorting code body, all SKUs that are not added to the corresponding picking batch are obtained to obtain an alternative SKU group.

[0067] Step 53 , based on the candidate SKU group, repeatedly perform the operations of indexing and selecting SKUs, adding the indexed and selected SKUs to the corresponding picking batches, and updating the active orders associated with the target seed wall until all SKUs in the third sorting code body are added to the corresponding picking batches.

[0068] In this embodiment, after all SKUs are picked to the end of the third sorting code body, it is determined whether there are any SKUs that were skipped before and have not been added to the corresponding picking batch. If so, index picking will be re-performed starting from the first position of the third sorting code body. During the re-index picking process, the SKUs that have been arranged to the corresponding picking batch will no longer be picked repeatedly, that is, by index picking in the SKU group consisting of all SKUs that have not been added to the corresponding picking batch. It can be understood that when the currently picked SKU is at the end of the third sorting code body, the current picking batch will be terminated, and when a new picking batch is started, the picking order of the picking batch will be considered, that is, whether to add the SKU for restarting index picking to the corresponding picking batch. Priority will be given to the picking order to which the last indexed SKU picked before restarting index picking belongs, and the target seed wall will give priority to maintaining the picking order to which the last indexed SKU picked belongs as the active order.

[0069] In this embodiment, the operations of indexing and picking SKUs, adding the indexed and picked SKUs to the corresponding picking batches, and updating the active orders associated with the target seed wall are repeated, and the repeated operations are continued until all SKUs of the corresponding third sorting encoding body are indexed and picked and batched to the corresponding picking batch positions.

[0070] In some optional implementations, all SKUs of the third sorting code body are sorted and batched by the following steps:

[0071] Step 1: Initialize the active order list active_orders=[] and the total picking batch set tours=[].

[0072] In this embodiment, active orders represents the list of activated orders on the target seed wall, which is used to store the picking orders that are currently partially picked but not yet fully picked. It will be updated while continuously allocating SKUs; tours represents the total picking batch, which is also the final output result; i represents the index position of the indexed picked SKU in the third sorting code body; PickCap and PutCap are the initial input values, representing the picking equipment capacity and the seed wall order capacity respectively; tour represents one of the picking batches. When specific conditions are met, such as the picking equipment is overloaded and cannot continue to put in SKUs, or i indexes to the end of the third sorting code body, the current batch will be terminated and stored in tours, and a new tour will be started to store the next batch of SKUs until all SKUs are allocated to the picking batch.

[0073] Step 2: Execute the main loop and gradually build the picking batch (execute the loop until all SKUs of all orders are picked), create a new picking batch tour=[] to store the SKUs of the current batch; initialize the picking batch load load=0, indicating that no SKUs are currently added; traverse i=1 starting from the first SKU of the third sorting code body.

[0074] Step 3: Enter the sub-loop, read the SKUs one by one, and build the picking batch.

[0075] This loop is executed continuously until the end of the third sorting code body is reached or the picking object load reaches the upper limit PickCap. The active order list is continuously updated in this loop.

[0076] Step 4: Complete the current picking batch and update the total picking batch set; recreate a new tour and continue the loop until all order SKUs are fully picked.

[0077] Step 5: Process the SKUs that have not been batched. If there are still SKUs that have not been included in any picking batch after the first traversal, a second traversal will be performed to process only the remaining SKUs until all SKUs are included in a picking batch.

[0078] Step 6: Output the final total picking batch plan and return the final tours.

[0079] In some optional implementations, there are only two order capacities of 2, and the orders to be picked are set as follows: Picking Order 1 (with SKU numbers: 1, 2, 3), Picking Order 2 (with SKU numbers: 4, 5, 6), Picking Order 3 (with SKU numbers: 7, 8), Picking Order 4 (with SKU numbers: 9, 10, 11, 12), Picking Order 5 (with SKU numbers: 13, 14), Picking Order 6 (with SKU numbers: 15, 16), Picking Order 7 ( SKU numbers are: 17, 18, 19, 20), picking order 8 (SKU numbers are: 21, 22, 23), picking order 9 (SKU numbers are: 24, 25, 26), picking order 10 (SKU numbers are: 27, 28). In the generated sorting code body, the picking batch process is performed. If the first index picked is SKU number 10, then the picking order to which SKU number 10 belongs (corresponding to picking order 4) will be activated first, and the picking order 10 will be SKU 10 is assigned to the first picking batch. When the next SKU is numbered 1, it will first be determined whether the picking order to which this SKU belongs has been activated. If not and there is still 1 order capacity (a new active order can be activated or associated), the picking order to which this SKU belongs (corresponding to picking order 1) will also be activated. The next SKU is numbered 11. The picking order to which it belongs is the same as the order to which SKU 10 belongs. If the picking equipment capacity allows (at this time there are only two SKUs in the picking batch and the picking equipment capacity is 4 SKUs), it can be assigned to the current picking batch. When the next SKU is numbered 27, it is found that it does not belong to the previously activated picking order, and two picking orders have already been activated (two active orders exist). In this case, it cannot be assigned to the current picking batch. The SKU of the two activated picking orders must be found later. For example, the index picks SKU numbered 9. The picking order to which it belongs is an activated order, so it can be assigned to the current batch.

[0080] In some embodiments, determining the fitness of the corresponding sequence encoding body is achieved by the following steps:

[0081] Step 61, determine all picking batches corresponding to the fourth sorting code body that has completed the current picking and sorting process, and search the storage location parameters corresponding to the SKU of each picking batch from the preset order storage location parameter information, wherein the fourth sorting code body includes one of the second sorting code body, the alternative sorting code body and the target sorting code body.

[0082] Step 62, using the preset batch picking time function and the corresponding storage location parameters, calculate the first picking time corresponding to each picking batch, and calculate the total picking time corresponding to the fourth sorting code body by minimizing the preset total picking time function and the corresponding first picking time.

[0083] In this embodiment, the following formula is used to calculate the first picking time corresponding to the picking batch: : .

[0084] In this embodiment, the corresponding picking time is calculated for the obtained picking batches, that is, the sum of the picking times of all picking batches is calculated. The picking time starts from the starting point (seeding wall), goes to the warehouse to pick according to the picking order of the picking batch, and finally returns to the starting point (seeding wall). According to the corresponding picking path, a picking distance can be obtained, and then the picking distance is divided by a set picking speed (the walking speed of the picking object) to obtain the picking time.

[0085] The corresponding total picking time is calculated using the following formula: ; The following formula is used to calculate the corresponding total order sorting completion time: ; The corresponding minimum picking manual work equilibrium parameters are calculated using the following formula: ; Where j represents the jth picking order, j∈{1,2,…,J}, J represents the picking order set, s i Indicates the sth i SKUs, i ∈{1,2,…,S}, S represents the SKU set, b represents the bth picking batch, b∈{1,2,…,B}, B represents the picking batch set; D represents the target seed wall location, c s1,s2 represents the shortest path length between the s1th SKU and the s2th SKU, which is a known value, c s represents the path length between the location of the s-th SKU and the target planting wall, V represents the walking speed of the picker, represents the picking time of the bth picking batch, Indicates the possible waiting time of the bth picking batch in the sorting process, Indicates the picking order of SKUs in the bth picking batch; Indicates whether the sth SKU of the jth picking order is the first picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the sth SKU of the jth picking order is the last picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the jth picking order is partially or completely picked in the bth picking batch, and if so, =1, otherwise, =0; Indicates the order of activity order associations on the target putwall, Indicates whether the jth picking order contains the sth SKU. If so, =1,q j represents the number of items (corresponding to SKUs) contained in the jth picking order, q b Indicates the number of items (corresponding to SKUs) contained in the bth picking batch.

[0086] In this embodiment, before calculating the fitness according to the corresponding formula, the following mathematical model is also established: 1. The capacity of the seeding wall is limited, and it is impossible to associate all picking orders at once; 2. The grid on the seeding wall is only associated with one picking order, and the next picking order can only be associated after all the items (SKUs) of the picking order are picked; 3. The capacity of each grid in the seeding wall is large enough; 4. The picking order information is known in advance, and each picking order contains at least one SKU; 5. Picking orders are allowed to be split; 6. After the picking batch is completed, the time for placing the items into the corresponding grid of the seeding wall is not considered; 6. There is no out-of-stock situation for the items on the shelf; 7. The number of items in the picking batch does not exceed the capacity of the picking equipment; 8. The picking object knows the storage location of the items in the warehouse; 8. The walking speed of the picking object remains constant; 9. The picking object starts picking from the seeding wall and returns to the seeding wall after completing the picking.

[0087] In this embodiment, the following constraints are established: 1. Association order constraints for orders on the seed wall For any two orders j1 and j2, there can only be a unique priority relationship in the grid association sorting on the seed wall. The constraint is: ; If order j1 is sorted before order j2, and order j2 is sorted before order j3, then order j1 is sorted before order j3, and the constraint is: ; If two orders have the same predecessor or successor, they should also have a priority relationship. Specifically, if order j1 and order j2 both have the same successor order j3, the priority of order j1 and order j2 must be guaranteed in the sorting. Similarly, if order j1 and order j2 have the same predecessor order j0, a priority relationship must also be established between them. The constraint is: ; .

[0088] 2. Constraints on picking batches ; The above constraint expresses the minimum number of priority ties required between orders based on putwall capacity, where mod is the remainder operator.

[0089] The constraint formula of picking equipment capacity constraint is: .

[0090] The picking path integrity constraint is: .

[0091] The picking path starts and ends at the put wall: .

[0092] Each picking path must start from the starting point at most once: .

[0093] 3. Order batching constraints Constraints that all SKUs in each picking order are assigned to picking batches: , , .

[0094] SKUs of two picking orders subject to priority constraints cannot be assigned to the same picking batch: , .

[0095] An order is either partially picked or fully picked in a batch, provided that at least one SKU associated with the order is selected in the batch. That is, if the SKU of an order has been assigned to a batch, the order must be fulfilled in that batch. The constraint is: , ; , .

[0096] If one picking order takes precedence over another picking order, then the two picking orders cannot be in the same picking batch. The constraint is: , .

[0097] The number of orders picked in each picking batch is limited to the maximum capacity of the put wall, .

[0098] Subpath elimination constraints: .

[0099] Step 63 , based on the total picking time, the preset order sorting completion time function and the picking batch picking completion time standard deviation function are used in sequence to calculate the total order sorting completion time and the minimum picking manual labor balance parameter corresponding to the fourth sorting code body.

[0100] Step 64 : Calculate the fitness corresponding to the fourth sorting code body based on the total picking time, the total order sorting completion time, and the balance parameter for minimizing picking manual work.

[0101] In some embodiments, before selecting the target sequence encoding body, the following steps are further performed:

[0102] Step 71: Determine whether the fitness of the sorted code body corresponding to the current iteration is greater than a preset fitness threshold.

[0103] Step 72: When it is determined that the fitness of the corresponding sorting code body is greater than the preset fitness threshold, the corresponding sorting code body is used as a candidate sorting code body.

[0104] Step 73 : Select the candidate sorting code body with the greatest fitness from the multiple candidate sorting code bodies to obtain the target sorting code body.

[0105] In some optional implementations, performing the non-dominated solution sorting and genetic evolution operations on the fifth sorted encoding body currently to be subjected to the non-dominated solution sorting and genetic evolution operations includes:

[0106] Step 1: Use a non-dominated sorting algorithm to perform non-dominated sorting on all fifth-ordered code bodies, and calculate the congestion degree of all fifth-ordered code bodies that have completed the non-dominated sorting to obtain the coding parameters corresponding to each fifth-ordered code body, wherein the coding parameters include non-dominated sorting parameters and congestion degree.

[0107] Step 2: Select the first non-dominated code entity from all the fifth-ranked code entities according to the non-dominated sorting parameter and the congestion degree.

[0108] Step 3: Using the NSGA-II algorithm, perform a genetic evolution operation on all fifth-ranked coding bodies to generate a preset number of first sub-coding bodies. The genetic evolution operation includes one of the following: tournament selection, crossover, and mutation. The crossover probability corresponding to the crossover adopts an adaptive crossover probability generated by a preset gradient-decreasing logsig function, and the mutation probability corresponding to the mutation adopts an adaptive mutation probability generated by a preset gradient-decreasing logsig function.

[0109] Step 4: The sixth sorted coding body obtained by merging all the first non-dominated coding bodies and all the first sub-coding bodies is subjected to non-dominated solution sorting and crowding calculation. According to the non-dominated sorting parameters and crowding, sorted coding bodies with the same number as the fifth sorted coding body are screened out from all the sixth sorted coding bodies to obtain sorted coding bodies that have completed one non-dominated solution sorting and genetic evolution operation.

[0110] This embodiment also provides a picking and sorting decision-making device that considers seed wall capacity. This device is used to implement the above-mentioned embodiments and preferred implementations, and details already described will not be repeated. As used below, terms such as "module," "unit," and "subunit" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0111] Figure 3 is a structural block diagram of a picking and sorting decision-making device considering the sowing wall capacity according to an embodiment of the present application, such as Figure 3 As shown, the device includes an encoding module 31, a processing module 32 and a generating module 33:

[0112] The encoding module 31 is configured to, after acquiring a plurality of to-be-picked picking orders trapped in a picking order pool corresponding to a target warehouse, randomly encode and initialize the SKUs associated with the plurality of picking orders according to a preset encoding method, thereby generating a plurality of first sorting encoding bodies, wherein each picking order is associated with at least one SKU, and the first sorting encoding body includes a plurality of subcodes representing a picking order using gene bits, each subcode corresponding to a SKU;

[0113] The processing module 32 is coupled to the encoding module 31 and is configured to utilize a heap optimized simulated annealing algorithm to perform sorting and correction preprocessing on the subcodes corresponding to the plurality of first sorting encoding bodies to generate a plurality of second sorting encoding bodies, and based on the plurality of second sorting encoding bodies, repeatedly execute non-dominated solution sorting and genetic evolution operations using the NSGA-II algorithm and the fitness of the corresponding sorting encoding bodies until a plurality of candidate sorting encoding bodies are generated, wherein the fitness is determined based on the objective function values ​​corresponding to the plurality of picking batches corresponding to the corresponding sorting encoding bodies, the picking batches being generated by picking and sorting the SKUs of the corresponding sorting encoding bodies according to a picking and sorting decision rule that takes into account the seed wall capacity, and the genetic evolution operation utilizes an adaptive crossover probability and an adaptive mutation probability;

[0114] The generation module 33 is coupled to the processing module 32 and is used to select a target sorting code body from multiple alternative sorting code bodies according to the fitness of the alternative sorting code bodies, and use the multiple target picking batches obtained by processing the target sorting code body according to the picking and sorting decision rules, and the picking path mapped by the sorting of all SKUs in each target picking batch as the decision result.

[0115] In some embodiments, the processing module 32 is further configured to, before determining the fitness of the corresponding sorting encoding body, Determine the picking capacity corresponding to the active order currently associated with the target seed wall and each picking batch to be constructed, wherein the active order is used to represent the picking order allowed to be loaded on the target seed wall, and the picking capacity is used to represent the number of SKUs picked by the picking equipment associated with the target seed wall for a picking batch; perform index picking on all SKUs of the third sorting code body currently to be picked and sorted, and determine whether the currently picked SKU belongs to the SKU of the currently associated active order; if it is determined that the currently picked SKU belongs to the SKU of the currently associated active order, add the currently picked SKU to an unfilled picking batch; repeat the steps of index picking the corresponding SKU, determining whether the picked SKU is the SKU corresponding to the active order, and adding the SKU to the corresponding picking batch until multiple picking batches corresponding to the currently associated active order are determined, and combine the multiple picking batches determined by the multiple decisions into a total picking batch, wherein the multiple picking batches corresponding to the third sorting code body include the total picking batch, and the number of SKUs added to each picking batch does not exceed the picking capacity.

[0116] In some embodiments, the processing module 32 is further configured to, when it is determined that the currently picked SKU does not belong to the SKU of the currently associated active order, determine whether the target seed wall allows the addition of associated active orders based on the number of currently associated active orders and the order loading capacity corresponding to the target seed wall; when it is determined that the target seed wall allows the addition of associated active orders, use the picking order to which the currently picked SKU belongs as the added active order, use the associated active orders and all added active orders as the currently associated active orders of the target seed wall, and add the currently picked SKU to an unfilled picking batch; when it is determined that the target seed wall does not allow the addition of associated active orders, perform index picking on the next SKU.

[0117] In some embodiments, the processing module 32 is also used to index and pick the next SKU after determining multiple picking batches corresponding to the currently associated active order, and determine the picking order to which the next SKU belongs; release all currently associated active orders, and associate at least the picking order to which the next SKU belongs as the corresponding active order, and add the next SKU picked by the index to a newly constructed picking batch.

[0118] In some embodiments, the processing module 32 is also used to determine whether the currently picked SKU is located at the end of the third sorting code body before combining multiple picking batches determined by multiple decisions into a total picking batch; when it is determined that the currently picked SKU is located at the end of the third sorting code body, all SKUs that have not been added to the corresponding picking batch are obtained to obtain an alternative SKU group; based on the alternative SKU group, the operations of indexing the picking SKU, adding the indexed picked SKU to the corresponding picking batch, and updating the active orders associated with the target seed wall are repeated until all SKUs of the third sorting code body are added to the corresponding picking batch.

[0119] In some embodiments, the processing module 32 is also used to determine all picking batches corresponding to the fourth sorting code body that completes the current picking and sorting processing, and to find the storage location parameters corresponding to the SKU of each picking batch from the preset order storage location parameter information, wherein the fourth sorting code body includes one of the second sorting code body, the alternative sorting code body and the target sorting code body; using the preset batch picking time function and the corresponding storage location parameters, the first picking time corresponding to each picking batch is calculated, and the total picking time corresponding to the fourth sorting code body is calculated by using the preset minimization total picking time function and the corresponding first picking time; according to the total picking time, the preset order sorting completion time function and the picking batch picking completion time standard deviation function are used in turn to calculate the order sorting total completion time and the minimization picking manual operation balance parameter corresponding to the fourth sorting code body respectively; based on the total picking time, the order sorting total completion time and the minimization picking manual operation balance parameter, the fitness corresponding to the fourth sorting code body is calculated.

[0120] In some embodiments, the processing module 32 calculates the first picking time corresponding to the picking batch using the following formula: : ; The corresponding total picking time is calculated using the following formula: ; The following formula is used to calculate the corresponding total order sorting completion time: ; The corresponding minimum picking manual work equilibrium parameters are calculated using the following formula: ; Where j represents the jth picking order, j∈{1,2,…,J}, J represents the picking order set, s i Indicates the sth i SKUs, i ∈{1,2,…,S}, S represents the SKU set, b represents the bth picking batch, b∈{1,2,…,B}, B represents the picking batch set; D represents the target seed wall location, c s1,s2 represents the shortest path length between the s1th SKU and the s2th SKU, which is a known value, c s represents the path length between the location of the s-th SKU and the target planting wall, V represents the walking speed of the picker, represents the picking time of the bth picking batch, Indicates the possible waiting time of the bth picking batch in the sorting process, Indicates the picking order of SKUs in the bth picking batch; Indicates whether the sth SKU of the jth picking order is the first picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the sth SKU of the jth picking order is the last picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the jth picking order is partially or completely picked in the bth picking batch, and if so, =1, otherwise, =0.

[0121] In some embodiments, the generation module 33 is also used to determine whether the fitness corresponding to the sorting code body after completing the current iteration is greater than a preset fitness threshold before selecting the target sorting code body; when it is determined that the fitness corresponding to the corresponding sorting code body is greater than the preset fitness threshold, the corresponding sorting code body is used as an alternative sorting code body; from multiple alternative sorting code bodies, the alternative sorting code body with the largest fitness is selected to obtain the target sorting code body.

[0122] This embodiment further provides a service platform, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0123] Optionally, the service platform may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0124] Optionally, in this embodiment, the processor may be configured to execute the following steps through a computer program:

[0125] S1, after obtaining multiple picking orders to be picked in the picking order pool corresponding to the target warehouse, randomly encode and initialize the population of the SKUs associated with the multiple picking orders according to a preset encoding method to generate multiple first sorting code bodies.

[0126] S2, using the heap optimized simulated annealing algorithm, performs sorting correction preprocessing on the subcodes corresponding to the multiple first sorting coding bodies to generate multiple second sorting coding bodies, and based on the multiple second sorting coding bodies, repeatedly executes the NSGA-Ⅱ algorithm and the fitness of the corresponding sorting coding bodies to perform non-dominated solution sorting and genetic evolution operation iterations until multiple alternative sorting coding bodies are generated.

[0127] S3, according to the fitness of the alternative sorting code body, select the target sorting code body from multiple alternative sorting code bodies, and process the target sorting code body according to the picking and sorting decision rules to obtain multiple target picking batches, and the picking path mapped by the sorting of all SKUs in each target picking batch is used as the decision result.

[0128] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be repeated here.

[0129] In addition, in conjunction with the above-described methods for picking and sorting decisions that consider seedwall capacity, embodiments of the present application may provide a storage medium for implementation. The storage medium stores a computer program that, when executed by a processor, implements any of the above-described methods for picking and sorting decisions that consider seedwall capacity.

[0130] The above embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A picking and sorting decision method considering the capacity of the planting wall, characterized in that: include: After obtaining a plurality of to-be-picked picking orders trapped in a picking order pool corresponding to a target warehouse, randomly encoding and population initializing the SKUs associated with the plurality of picking orders according to a preset encoding method is performed to generate a plurality of first sorting encoding bodies, wherein each of the picking orders is associated with at least one SKU, and the first sorting encoding body includes a plurality of subcodes representing a picking order using gene bits, and each subcode corresponds to one SKU; Using a heap-optimized simulated annealing algorithm, the subcodes corresponding to the plurality of first sorting code bodies are subjected to sorting correction preprocessing to generate a plurality of second sorting code bodies. Based on the plurality of second sorting code bodies, the NSGA-II algorithm and the fitness of the corresponding sorting code bodies are repeatedly executed to perform non-dominated solution sorting and genetic evolution operations iteratively until a plurality of candidate sorting code bodies are generated, wherein the fitness is determined according to the objective function values ​​corresponding to the plurality of picking batches corresponding to the corresponding sorting code bodies, the picking batches being generated by picking and sorting the SKUs of the corresponding sorting code bodies according to a picking and sorting decision rule that takes into account the seed wall capacity, and the genetic evolution operation adopts an adaptive crossover probability and an adaptive mutation probability; According to the fitness of the alternative sorting code body, a target sorting code body is selected from multiple alternative sorting code bodies, and multiple target picking batches obtained by processing the target sorting code body according to the picking and sorting decision rules, and the picking path mapped by the sorting of all the SKUs in each target picking batch are used as the decision result.

2. The method according to claim 1, characterized in that Before determining the fitness of the corresponding sorting encoding body, the method further includes: Determine the active order currently associated with the target seed wall and the picking capacity corresponding to each of the picking batches to be constructed, wherein the active order is used to represent the picking order allowed to be loaded on the target seed wall, and the picking capacity is used to represent the number of SKUs picked by the picking equipment associated with the target seed wall for one picking batch; Perform index picking on all the SKUs of the third sorting code body currently to be picked and sorted, and determine whether the currently picked SKU belongs to the SKU of the currently associated active order; When it is determined that the currently picked SKU belongs to the SKU of the currently associated active order, the currently picked SKU is added to an unfilled picking batch; Repeat the steps of index picking the corresponding SKU, determining whether the picked SKU is the SKU corresponding to the active order, and adding the SKU to the corresponding picking batch, until a plurality of picking batches corresponding to the currently associated active order are decided, and combine the plurality of picking batches decided separately into a total picking batch, wherein the plurality of picking batches corresponding to the third sorting code body include the total picking batch, and the number of the SKUs added in each picking batch does not exceed the picking capacity.

3. The method according to claim 2, characterized in that In the case where it is determined that the currently picked SKU does not belong to the SKU of the currently associated active order, the method further includes: Determining whether the target seedwall allows additional associated active orders based on the number of currently associated active orders and the order loading capacity corresponding to the target seedwall; If it is determined that the target seedwall allows adding the associated active order, the picking order to which the currently picked SKU belongs is used as the added active order, and the associated active order and all the added active orders are used as the currently associated active orders of the target seedwall, and the currently picked SKU is added to an unfilled picking batch; When it is determined that the target seed wall does not allow the associated active order to be added, index picking is performed on the next SKU.

4. The method according to claim 3, characterized in that After determining the plurality of picking batches corresponding to the currently associated active order, the method further includes: Index and pick the next SKU, and determine the picking order to which the next SKU belongs; All currently associated active orders are released, and at least the picking order to which the next SKU belongs is associated as the corresponding active order, and the next SKU picked by the index is added to a newly constructed picking batch.

5. The method according to claim 4, characterized in that Before combining the plurality of picking batches determined by multiple decisions into a total picking batch, the method further includes: Determine whether the currently selected SKU is at the end of the third sorting code body; When it is determined that the currently picked SKU is at the end of the third sorting code body, all SKUs that are not added to the corresponding picking batch are obtained to obtain an alternative SKU group; Based on the alternative SKU group, the operations of index picking the SKU, adding the index picked SKU to the corresponding picking batch, and updating the active order associated with the target seed wall are repeatedly performed until all the SKUs of the third sorting code body are added to the corresponding picking batch.

6. The method according to claim 1, characterized in that Determining the fitness of the corresponding sorting encoding body includes: Determine all the picking batches corresponding to the fourth sorting code body that has completed the current picking and sorting process, and search for the storage location parameters corresponding to the SKU of each picking batch from the preset order storage location parameter information, wherein the fourth sorting code body includes one of the second sorting code body, the candidate sorting code body, and the target sorting code body; Calculate the first picking time corresponding to each picking batch using a preset batch picking time function and the corresponding storage location parameters, and calculate the total picking time corresponding to the fourth sorting code body using a preset minimization total picking time function and the corresponding first picking time; According to the total picking time, the preset order sorting completion time function and the picking batch picking completion time standard deviation function are used in sequence to calculate the total order sorting completion time and the minimum picking manual work balance parameter corresponding to the fourth sorting code body; The fitness corresponding to the fourth sorting encoding body is calculated based on the total picking time, the total order sorting completion time and the minimization of picking manual work balance parameter.

7. The method according to claim 6, characterized in that The following formula is used to calculate the first picking time corresponding to the picking batch: : ； The corresponding total picking time is calculated using the following formula: ; The following formula is used to calculate the corresponding total order sorting completion time: ； The corresponding minimum picking manual work equilibrium parameters are calculated using the following formula: ； Where j represents the jth picking order, j∈{1,2,…,J}, J represents the picking order set, s i Indicates the sth i SKUs, i ∈{1,2,…,S}, S represents the SKU set, b represents the bth picking batch, b∈{1,2,…,B}, B represents the picking batch set; D represents the target seed wall location, c s1,s2 represents the shortest path length between the s1th SKU and the s2th SKU, which is a known value, c s represents the path length between the location of the s-th SKU and the target planting wall, V represents the walking speed of the picker, represents the picking time of the bth picking batch, Indicates the possible waiting time of the bth picking batch in the sorting process, Indicates the picking order of SKUs in the bth picking batch; Indicates whether the sth SKU of the jth picking order is the first picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the sth SKU of the jth picking order is the last picked SKU in the bth picking batch. If so, =1, otherwise, =0, Indicates whether the jth picking order is partially or completely picked in the bth picking batch, and if so, =1, otherwise, =0.

8. The method according to claim 1, characterized in that Before selecting the target sequence encoding body, the method further includes: Determine whether the fitness corresponding to the sorted encoding body after the current iteration is completed is greater than a preset fitness threshold; In the case where it is determined that the fitness corresponding to the corresponding sorting code body is greater than a preset fitness threshold, the corresponding sorting code body is used as the candidate sorting code body; From the multiple candidate sorting code bodies, the candidate sorting code body with the largest fitness is selected to obtain the target sorting code body.

9. A service platform comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps of the picking and sorting decision method considering the sowing wall capacity according to any one of claims 1 to 8.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the picking and sorting decision method considering the seed wall capacity according to any one of claims 1 to 8 is implemented.

Images