Method applied to warehouse replenishment and storage location collaborative planning

Abstract

A method applied to warehouse replenishment and storage location collaborative planning comprises the steps that a warehouse replenishment and storage location collaborative model is constructed, and the target of the warehouse replenishment and storage location collaborative model is that the total energy consumption of warehouse carrying equipment is minimum and the warehouse channel load is minimum; and solving the warehouse replenishment and storage location collaboration model based on an improved non-dominated sorting elite genetic algorithm (ENSGA-II) to obtain an optimal population solution and a corresponding lowest energy consumption value and channel use difference coefficient. According to the method, Pareto leading edge grade sorting is performed on the obtained solutions, and an elitist solution is selected based on a knee region point criterion so as to find an approximate optimal solution which is excellent in performance in multiple target dimensions. Compared with a multi-objective evolutionary algorithm based on indexes and decomposition, the method has the optimal theoretical support of a non-dominated sorting criterion and a knee region point criterion, and the search performance and stability of the algorithm are ensured.

Description

Technical Field

[0001] The present invention relates to the field of intelligent warehousing technology, and in particular to a method for warehouse replenishment and storage space collaborative planning. Background Art

[0002] With the opening of the prelude of intelligent manufacturing 5.0, in order to meet customers' demand for small batches, multiple batches and short time windows for food and medicine, the intelligent logistics system represented by the food and medicine storage and picking system has gradually developed into an operation mode combining the storage area and the forward area, and the picking efficiency of the storage system has achieved a qualitative leap. Scholars define such a warehouse as a forward-reserve warehouse (FRW). The storage area is composed of an automated storage and retrieval system (AS / RS) picking system, and the forward area is composed of a robotic mobile fulfillment system (RMFS). The storage area provides replenishment for the forward area. The operation process of the warehouse system is as follows: 1) The stacker transports the pallet with SKU to the replenishment station at the exit. 2) The automatic guided vehicle AGV in the forward area transports the mobile shelves that need to be replenished to the replenishment station. 3) The picker puts the SKU into the mobile shelf according to the replenishment demand quantity. 4) AGV transports the replenished mobile shelves to the front area storage. 5) According to the order requirements, AGV transports the mobile shelves to the picking platform to wait for picking. 6) The pickers pick the SKUs, and the AGV transports the mobile shelves that have been picked to the front area for storage. It can be seen that the replenishment decision and the commodity storage location allocation decision of the front-storage warehouse FRW occur at the same time, and the collision and deadlock of the transport equipment will be an important factor affecting the picking efficiency of the system. Appropriate storage location allocation planning will effectively avoid the collision and deadlock of the transport equipment from the source.

[0003] However, with the development of smart warehouses, existing research technologies can no longer meet the needs of actual system operation goals. First, most research technologies only explore storage location allocation problems / batch picking problems / path planning problems in warehouses, and lack research on related replenishment-storage location collaborative planning technologies. Secondly, most researchers focus on the storage location allocation problem with economic efficiency as a single goal, and lack multi-objective collaborative optimization technology that takes into account both economic and green goals. Finally, the extensive use of automatic guided vehicles and stackers in front-storage warehouses has brought congestion problems, and there is a lack of research on related obstacle avoidance technologies. Therefore, there is an urgent need for a method applied to warehouse replenishment and storage location collaborative planning to solve this problem. Summary of the invention

[0004] In view of the above problems, the present invention is proposed to provide a method for warehouse replenishment and storage location collaborative planning that overcomes the above problems or at least partially solves the above problems.

[0005] In order to solve the above technical problems, the embodiments of the present application disclose the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention discloses a method for warehouse replenishment and storage location collaborative planning, comprising:

[0007] S100. Construct a warehouse replenishment and storage coordination model, the goal of which is to minimize the total energy consumption of warehouse transportation equipment and the warehouse channel load; the decision variable in the warehouse replenishment and storage coordination model is the variable y of the sub-storage k of shelf l assigned to channel r by SKUp pklr SKUp is assigned to the front channel r 1 The variable of the sub-storage position i on SKUp is assigned to the front channel r 2 The variable of the sub-slot j on SKU p in the front area channel r 1 Number of picks on SKU p in store aisle r 2 Number of picks on

[0008] S200. Solve the warehouse replenishment and storage coordination model based on the improved non-dominated sorting elite genetic algorithm, and obtain the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient based on the knee point criterion.

[0009] Furthermore, in S100, the minimum total energy consumption function of the warehouse transportation equipment is:

[0010]

[0011] Where R represents the channel set, r = 1, ... |R|, r∈R; r 1 ∈R 1 Indicates the channel index of the front area, r 2 ∈R 2 Represents the channel index of the storage area; I represents the sub-storage location set of each channel in the front area, i=1,...|I|, i∈I; J represents the sub-storage location set of each channel in the storage area, j=1,...|J|, j∈J; P represents the SKU set, p=1,...|P|, p∈P; Indicates that the AGVs transport SKU p from the front area r 1 The energy consumption required to get from storage location i to the picking station, Indicates the picking frequency of SKUp in each sub-storage location in the front area. Indicates that SKUp is assigned to the front channel r 1 The sub-storage position i on the top; Indicates that the robot moves SKUp from storage area r 2 The energy consumption required to move from storage location j to the front-end picking station; represents the picking frequency of SKU p in each sub-storage location in the storage area, Indicates that SKU p is assigned to storage area channel r 2 The sub-storage position j on the top; Indicates that the robot moves SKUp from storage area r 2 The storage position j to the front area r 1 The energy consumption required for storage position i; It represents the average replenishment frequency of each sub-storage location assigned to SKU p from the picking station to the front area. It indicates the average replenishment frequency of each sub-storage location allocated by SKUp from the storage area to the front area to the picking station.

[0012] Furthermore, in S100, the warehouse channel load minimum function is:

[0013]

[0014] in, For SKUp in the front channel 1 The number of picks on is the average number of picking times in all channels of the front area, For SKU in store aisle 2 The number of picks on is the average number of picks across all aisles in the storage area.

[0015] Furthermore, in S100, the constraints of the minimum total energy consumption function of the warehouse transportation equipment and the minimum warehouse channel load function include at least: one sub-storage location can only store one SKU, the number of SKU types stored on a single shelf cannot exceed the total sub-storage locations of a single shelf, and the number of sub-storage locations occupied by each SKU cannot exceed the total allocated storage locations.

[0016] Further, in S200, the warehouse replenishment and storage location coordination model is solved based on an improved non-dominated sorting elite genetic algorithm to obtain an optimal population solution and a corresponding minimum energy consumption value and channel usage difference coefficient. The specific method includes:

[0017] S201. Set parameters and initialize the population;

[0018] S202. Generating a selected population based on the initialization population;

[0019] S203. Generate a crossover population based on the crossover method;

[0020] S204. Correcting the crossover population using a correction method for an infeasible solution to obtain a corrected population;

[0021] S205. mutating the corrected population using a mutation method to obtain a mutated population;

[0022] S206. Merge the initial population and the mutated population to form a merged population, perform non-dominated sorting on the merged population, and obtain the Pareto rank of each solution in the merged population;

[0023] S207. Based on the Pareto level of each solution in the merged population, a new offspring is generated, the number of iterations is updated, and it is determined whether the current number of iterations reaches the maximum number of iterations. If the maximum number of iterations is reached, the process goes to S208; if the maximum number of iterations is not reached, the process returns to S202;

[0024] S208. The current number of iterations reaches the maximum number of iterations, and the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient are obtained.

[0025] Furthermore, in S201, the parameters are set and the population is initialized. The specific method includes: setting the parameters, the parameters including at least the total number of storage spaces, the number of empty storage spaces and the number of non-empty storage spaces in the front area and the storage area, and the number of sub-storage spaces allocated to each SKU; determining the initial solution population size according to the parameters, randomly selecting a SKU and placing it in the sub-storage spaces in the front area and the storage area until the number of sub-storage spaces allocated to the SKU is allocated, and then allocating storage spaces to the next SKU, and repeating the above steps until storage space allocation is completed for all SKUs.

[0026] Further, in S202, a selection population is generated based on the initialization population, and the specific method includes: performing non-dominated sorting on the initialization population to obtain a Pareto front set; judging the merits of the solution according to the merits of the Pareto front set, and selecting elite individuals using a binary tournament technique until the elite individuals reach a preset solution population size, at which time the population composed of elite individuals is the selection population.

[0027] Further, in S203, a crossover population is generated based on the crossover method, and the crossover method includes a crossover method based on an optimal target and a two-point crossover method; wherein the crossover method based on an optimal target specifically includes: randomly selecting the optimal chromosome S on two target dimensions based on the selected population best1 and S best2 , respectively select the part between the preamble area and the storage area and a parent chromosome Parent 1 Cross the corresponding parts to generate new offspring 1 The two-point crossover method specifically includes: randomly selecting genes between two parent chromosomes for exchange, and obtaining two new offspring chromosomes. 2 , offspring 3 ; Until all traversals complete all selected population solutions, that is, a crossover population is generated.

[0028] Furthermore, in S204, the crossover population is corrected by using a correction method for an infeasible solution to obtain a corrected population, and the specific method includes: checking in turn the number of sub-storage positions occupied by each SKU in the chromosome, eliminating redundant SKUs, and replacing SKUs with missing sub-storage positions, until the number of storage positions occupied by all SKUs meets the allocated number of sub-storage positions, at which point the model has a feasible solution.

[0029] Further, in S205, a mutation method is used to mutate the corrected population to obtain a mutated population, and the specific method includes: selecting a parent chromosome in the crossover population according to a preset mutation probability for a mutation operation, and randomly selecting genes in the pre-region and the storage region for exchange to complete the mutation operation, until all crossover population solutions are traversed and completed, that is, a mutated population is generated.

[0030] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:

[0031] The present invention discloses a method for warehouse replenishment and storage space collaborative planning, which searches for approximate optimal solutions that perform well in multiple objective dimensions by ranking the obtained solutions according to the Pareto front and selecting elite solutions based on the knee point criterion. Compared with multi-objective evolutionary algorithms based on indicators and decomposition, the present invention has the optimal theoretical support of non-dominated sorting criteria and knee point criteria, which ensures the search performance and stability of the algorithm. The present invention creatively proposes adaptive mutation technology to enhance the development capabilities of search individuals, thereby improving the accuracy of the optimal solution. The balance between local search and global search can not only improve the search efficiency of the algorithm, but also improve the search accuracy of the algorithm. The present invention uses elite solution selection criteria based on congestion and knee point criteria to reduce the difficulty of decision makers in making decisions due to the large number of Pareto front solutions in the later stages of iteration, thereby enhancing the practical applicability of the algorithm.

[0032] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0034] Figure 1 This is a flow chart of a method for warehouse replenishment and storage location collaborative planning in Example 1 of the present invention;

[0035] Figure 2 This is a schematic diagram of setting commodity information and storage location allocation parameters in a front-storage warehouse in Example 1 of the present invention.

[0036] Figure 3 This is a schematic diagram of a chromosome encoding method in Example 1 of the present invention;

[0037] Figure 4 Schematic diagram of a crossover method based on an optimal target in Embodiment 1 of the present invention;

[0038] Figure 5 This is a schematic diagram of a two-point intersection method in Embodiment 1 of the present invention;

[0039] Figure 6 Schematic diagram of a method for correcting an infeasible solution in Embodiment 1 of the present invention;

[0040] Figure 7 This is a schematic diagram of a variation method in Example 1 of the present invention;

[0041] Figure 8 This is a comparative analysis diagram of Pareto solution sets under different numbers of iterations in Example 2 of the present invention;

[0042] Fig. 9 It is a convergence trend diagram of ENSGA-II and NSGA-II in Example 2 of the present invention;

[0043] Fig.10 It is a schematic diagram of the optimal solution of ENSGA-II and the optimal solution of NSGA-II under the medium scale of s2 in Example 2 of the present invention;

[0044] Fig.11 It is a schematic diagram of the optimal solution of ENSGA-II and the optimal solution of NSGA-II under large scale of s3 in Example 2 of the present invention;

[0045] Fig.12 This is a schematic diagram of the optimal solution of ENSGA-II and the optimal solution of NSGA-II under the ultra-large scale of s4 in Example 2 of the present invention. DETAILED DESCRIPTION

[0046] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0047] Example 1

[0048] In order to solve the problems existing in the prior art, an embodiment of the present invention provides a method for warehouse replenishment and storage location collaborative planning, such as Figure 1 ,include:

[0049] S100. Construct a warehouse replenishment and storage coordination model, the goal of which is to minimize the total energy consumption of warehouse transportation equipment and the warehouse channel load; the decision variable in the warehouse replenishment and storage coordination model is the variable y of the sub-storage k of shelf l assigned to channel r by SKUp pklr SKUp is assigned to the front channel r 1 The variable of the sub-storage position i on SKUp is assigned to the front channel r 2 The variable of the sub-slot j on SKU p in the front area channel r 1 Number of picks on SKU p in store aisle r 2 Number of picks on

[0050] Specifically, we first introduce the multi-objective optimization problem applied to warehouse replenishment and storage space collaborative planning. Figure 2As shown in the figure, the front area is some low-level mobile shelf storage locations, and there are different sub-storage locations on the mobile shelves, each of which stores one SKU; the storage area is some fixed high-level shelves, each grid represents a storage location, and each storage location can only store one SKU. The storage area and the front area have different storage capacity. By minimizing the green target, i.e., energy consumption, and the efficiency target congestion degree, the dual-objective optimization is used to determine how many items should be placed in the storage area and the replenishment items in the front area, and in which shelf and storage layer of each system. By ensuring that the number of times each channel is used fluctuates smoothly, it is possible to avoid excessive congestion in a channel due to a large number of times of use, which leads to a decrease in the system picking efficiency. According to existing research, placing frequently purchased items on the same shelf can reduce energy consumption, and placing items with high purchase frequency in a storage location close to the picking platform can greatly reduce energy consumption. In addition, the product picking storage location can determine the picking channel of the transport equipment. Planning the replenishment-storage location allocation can reduce the congestion of the picking channel and improve the picking efficiency. Therefore, there are two key issues that need to be resolved in the replenishment-storage allocation synchronization problem in the forward storage warehouse of this embodiment: (1) the automatic storage and retrieval system (AS / AR system) in the storage area needs to decide which storage location in which channel the incoming goods should be placed, and (2) at the same time, the robotic mobile fulfillment system (RMFS) in the forward area needs to decide which goods need to be replenished and the replenishment quantity, as well as which shelf and layer the replenished goods should be placed on.

[0051] This embodiment takes into account picking efficiency and energy consumption and is modeled as a multi-objective optimization problem. The two objective functions are to minimize the total energy consumption of warehouse transportation equipment and to minimize the warehouse channel load; specifically, when constructing a warehouse replenishment and storage location coordination model, based on theory and reality, the model must ensure the following assumptions: (1) Only one SKU is stored in a storage location; (2) One SKU can be stored in multiple storage locations; (3) Vacant storage locations are allowed. Before constructing the warehouse replenishment and storage location coordination model, the model parameters and model variables are first defined through Tables 1-3. Specifically, Table 1 is a set of model parameters, Table 2 is a specific model parameter, and Table 3 is a model decision variable.

[0052] Table 1

[0053]

[0054]

[0055] Table 2

[0056]

[0057]

[0058] Table 3

[0059]

[0060]

[0061] In this embodiment, through the parameters in Tables 1-3, the minimum total energy consumption function of the warehouse transportation equipment is obtained as follows:

[0062]

[0063] Where R represents the channel set, r = 1, ... |R|, r∈R; r 1 ∈R 1 Indicates the channel index of the front area, r 2 ∈R 2 Represents the channel index of the storage area; I represents the sub-storage location set of each channel in the front area, i=1,...|I|, i∈I; J represents the sub-storage location set of each channel in the storage area, j=1,...|J|, j∈J; P represents the SKU set, p=1,...|P|, p∈P; Indicates that the AGVs transport SKU p from the front area r 1 The energy consumption required to get from storage location i to the picking station, Indicates the picking frequency of SKUp in each sub-storage location in the front area. Indicates that SKUp is assigned to the front channel r 1 The sub-storage position i on the top; Indicates that the robot moves SKUp from storage area r 2 The energy consumption required to move from storage location j to the front-end picking station; represents the picking frequency of SKU p in each sub-storage location in the storage area, Indicates that SKU p is assigned to storage area channel r 2 The sub-storage position j on the top; Indicates that the robot moves SKU p from storage area r 2 The storage position j to the front area r 1 The energy consumption required for storage position i; It indicates the average replenishment frequency of each sub-storage location allocated by SKUp from the picking station to the front area. It indicates the average replenishment frequency of SKUp from each sub-location allocated in the storage area to the picking station.

[0064] The warehouse channel load minimum function is:

[0065]

[0066] in, For SKUp in the front channel 1 The number of picks on is the average number of picking times in all channels of the front area, For SKU in store aisle r 2 The number of picks on is the average number of picks across all aisles in the storage area.

[0067] In this embodiment, in S100, the constraints of the minimum total energy consumption function of the warehouse transportation equipment and the minimum warehouse channel load function include at least: one sub-storage location can only store one SKU, the number of SKU types stored on a single shelf cannot exceed the total sub-storage locations of a single shelf, and the number of sub-storage locations occupied by each SKU cannot exceed the total allocated storage locations.

[0068] The specific constraint formula is:

[0069]

[0070] Among them, constraints (1)(2)(3)(4) indicate that a sub-storage location can only store one SKU. Constraints (5)(6)(7)(8) indicate that the number of SKU types stored on a single shelf cannot exceed the total number of sub-storage locations on a single shelf. Constraints (9)(10) ensure that the number of sub-storage locations occupied by each SKU cannot exceed the total number of allocated storage locations. Constraints (11)-(13) define the decision variables.

[0071] S200. Solve the warehouse replenishment and storage coordination model based on the improved non-dominated sorting elite genetic algorithm, and obtain the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient based on the knee point criterion.

[0072] In S200 of this embodiment, the warehouse replenishment and storage location coordination model is solved based on an improved non-dominated sorting elite genetic algorithm to obtain an optimal population solution and a corresponding minimum energy consumption value and channel usage difference coefficient. The specific method includes:

[0073] S201. Setting parameters and initializing the population; specifically, in S201 of the present embodiment, setting parameters and initializing the population, the specific method includes: setting parameters, the parameters at least including the total number of storage spaces in the front area and the storage area, the number of empty storage spaces and the number of non-empty storage spaces, and the number of sub-storage spaces allocated to each SKU; determining the initial solution population size according to the parameters, randomly selecting a SKU and placing it in the sub-storage spaces in the front area and the storage area until the number of sub-storage spaces allocated to the SKU is allocated, and then allocating storage spaces to the next SKU, and repeating the above steps until storage space allocation is completed for all SKUs.

[0074] For example, Figure 2 The total number of storage spaces, empty storage spaces and non-empty storage spaces in the pre-position area and storage area, as well as the number of sub-storage spaces allocated to each SKU are given. Based on the above parameters, the initial solution population of the present invention is set to 100. The initial solution chromosome is as follows Figure 3 The figure shows a 2*28 matrix, which is coded according to the total number of sub-storage positions in the front-storage warehouse. There are 28 sub-storage positions, i.e. 28 genes (2*1 matrix), the first 16 columns are the front area codes, and the last 12 columns are the storage area codes. The numbers in the first row of squares represent the SKU numbers, and the column where the SKU numbers are located is the number of sub-storage positions. "0" indicates an empty storage position. The second row indicates the number of SKUs stored in the corresponding sub-storage position.

[0075] S202. Generate a selection population based on the initialization population; in S202 of the present embodiment, generate a selection population based on the initialization population, and the specific method includes: performing non-dominated sorting on the initialization population to obtain a Pareto front set; judging the merits of the solution according to the merits of the Pareto front set, and selecting elite individuals using a binary tournament technique until the number of elite individuals reaches a preset solution population size, at which time the population composed of elite individuals is the selection population.

[0076] S203. Generate a crossover population based on the crossover method; in S203 of this embodiment, Figure 4 , generating a crossover population based on the crossover method, wherein the crossover method includes a crossover method based on an optimal target and a two-point crossover method; wherein the crossover method based on an optimal target specifically includes: randomly selecting the optimal chromosome S on two target dimensions based on the selected population best1 and S best2 , respectively select the part between the preamble area and the storage area and a parent chromosome Parent 1 Cross the corresponding parts to generate new offspring 1 ;like Figure 5 The two-point crossover method specifically includes: randomly selecting genes between two parent chromosomes for exchange, and obtaining two new offspring chromosomes. 2 , offspring 3 ; Until all traversals complete all selected population solutions, that is, a crossover population is generated.

[0077] S204. Use the correction method of the infeasible solution to correct the crossover population to obtain a corrected population; since the crossover technology changes the total number of sub-storage digits occupied by each SKU, a solution that violates the model constraints will be generated, which is called an infeasible solution. In order to make the infeasible solution a feasible solution, in S204 of this embodiment, Figure 6, the crossover population is corrected by using the correction method of the infeasible solution to obtain the corrected population. The specific method includes: checking the number of sub-storage positions occupied by each SKU in the chromosome in turn, eliminating the redundant SKUs, and replacing the SKUs with missing sub-storage positions until the number of storage positions occupied by all SKUs meets the allocated number of sub-storage positions, at which time the model has a feasible solution.

[0078] S205. Use a mutation method to mutate the modified population to obtain a mutated population; In S205 of this embodiment, a mutation method is used to mutate the modified population to obtain a mutated population. The specific method includes: selecting a parent chromosome in the crossover population according to a preset mutation probability to perform a mutation operation, and randomly selecting genes in the pre-region and the storage region to exchange and complete the mutation operation, until the generated new chromosome reaches the population size, that is, a mutated population is generated. Specifically, Figure 7 As shown, in the crossover population, a parent chromosome is selected for mutation operation according to the mutation probability of 0.1, and the genes of the pre-region Bit1 and Bit2, and the storage region Bit3 and Bit4 are randomly selected for exchange to complete the mutation operation. Since the mutation operation is a change made within the same chromosome, no infeasible solution will be generated. Until all traversals complete all crossover-corrected population solutions, a mutant population is generated. At this point, the entire evolution process is completed and enters S206.

[0079] S206. Merge the initial population and the mutated population to form a merged population, perform non-dominated sorting on the merged population, and obtain the Pareto rank of each solution in the merged population;

[0080] S207. Based on the Pareto level of each solution in the merged population, a new offspring is generated, the number of iterations is updated, and it is determined whether the current number of iterations reaches the maximum number of iterations. If it reaches the maximum number of iterations, the process goes to S208; if it does not reach the maximum number of iterations, the process returns to S202; specifically, excellent solutions are selected in turn according to the Pareto level of the solutions in the merged population to enter the offspring. If the levels are the same, the solution with the largest crowding distance is selected to enter the offspring. This is done until the population size is reached, i.e., a new offspring is generated.

[0081] S208. The current number of iterations reaches the maximum number of iterations, and the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient are obtained.

[0082] This embodiment discloses a method for warehouse replenishment and storage space collaborative planning, which searches for approximate optimal solutions that perform well in multiple objective dimensions by ranking the obtained solutions on the Pareto front and selecting elite solutions based on the knee point criterion. Compared with multi-objective evolutionary algorithms based on indicators and decomposition, the present invention has the Pareto optimal theoretical support of multi-objective evolutionary algorithms and economics, which ensures the search performance and stability of the algorithm. The present invention creatively proposes adaptive mutation technology to enhance the development capabilities of search individuals, thereby improving the accuracy of the optimal solution. The balance between local search and global search can not only improve the search efficiency of the algorithm, but also improve the search accuracy of the algorithm. The present invention is based on the optimal theoretical support of the non-dominated sorting criterion and the knee point criterion, which ensures the search performance and stability of the algorithm.

[0083] Example 2

[0084] In order to verify the effect of the non-dominated elite sorting genetic algorithm (ENSGA-II) disclosed in Example 1, a comparison scheme of the non-dominated elite sorting genetic algorithm (NSGA-II) is designed. Table 4 shows the basic data settings of the multi-objective model of replenishment-storage allocation collaborative planning. The crossover probability of the second generation of the non-dominated elite sorting genetic algorithm is 0.8, the mutation probability is 0.1, the population size is 50, and the maximum number of iterations is 50.

[0085] Table 4 Basic data range

[0086]

[0087]

[0088] In the embodiment, the above fixed scale setting will be used to analyze and compare the visualization results of the Pareto frontier solutions of the two algorithms at different iterations. Figure 8 As shown in the figure, these four figures describe the Pareto optimal solutions when the number of iterations is 10, 20, 30 and 50 respectively. The vertical axis represents the first target value (total energy consumption), and the horizontal axis represents the second target value (total balance factor). Since the minimum value of the objective function is sought, the closer the scattered point solution is to the lower right corner area, the higher the quality of the solution, and the wider the distribution of the scattered points on the horizontal and vertical axes, the better the diversity of solutions obtained by the algorithm. The red triangle represents the Pareto optimal solution obtained by ENSGA-II, and the blue circle represents the Pareto optimal solution obtained by NSGA-II. Figure 7 As shown in the figure, when the number of iterations is 10 / 20 / 30, the scatter points of the red triangle are closer to the lower left corner, and the optimal solution of ENSGA-II is better than the optimal solution of NSGA-II. When the number of iterations is 50, both algorithms obtain the optimal solution. Therefore, the above analysis results show that ENSGA-II can obtain a better Pareto optimal solution in a shorter time.

[0089] In this embodiment, we analyze in detail the convergence speed of the dual objective values ​​obtained by the two algorithms. Fig. 9 As shown in the figure, the convergence trends of ENSGA-II and NSGA-II are in the upper and lower rows, respectively. The left sub-figure is the convergence trend of the first target value (total energy consumption), and the right sub-figure is the convergence trend of the second target value (total channel balance factor). The vertical axis represents the ratio of the target value, and the horizontal axis represents the change in the number of iterations. Fig. 9 The results of a) show that the target value of total energy consumption reaches stability after 5 iterations, while the target value of total channel balance factor reaches stability after 3 iterations. The Pareto optimal solution obtained by the ENSGA-II algorithm reaches stability after 5 iterations. Fig. 9 As shown in b), after 11 iterations, NSGA-II can obtain the same total energy consumption target value as ENSGA-II and achieve stability, while the target value of the total channel balance factor requires 4 iterations. After 11 iterations, the Pareto optimal solution obtained by NSGA-II is stable. The above analysis results show that ENSGA-II can obtain a better Pareto optimal solution in a shorter time and achieve stability first.

[0090] Visual analysis of Pareto fronts of different scales under different iterations. In the embodiment, medium-scale, large-scale and ultra-large-scale examples are compared and analyzed to verify the performance of the algorithm. Figure 10-12 As shown in the figure, these four figures describe the Pareto optimal solutions when the number of iterations is 10, 20, 30 and 50 respectively. The vertical axis represents the first target value (total energy consumption), and the horizontal axis represents the second target value (total channel balance factor). Since the minimum value of the objective function is sought, the closer the scattered point solution is to the lower left corner area, the higher the quality of the solution, and the wider the distribution of the scattered points on the horizontal and vertical axes, the better the diversity of solutions obtained by the algorithm. The red triangle represents the Pareto optimal solution obtained by ENSGA-II, and the blue circle represents the Pareto optimal solution obtained by NSGA-II. Figure 8As shown in the figure, under the medium scale of s2, no matter how many iterations, the red triangle is closer to the lower right corner area (high solution quality) and widely distributed (good diversity), so the optimal solution of ENSGA-II is significantly better than the optimal solution of NSGA-II. Under the large scale of s3 and the super large scale of s4, when the number of iterations is 30 / 50, the red triangle is closer to the lower right corner area (high solution quality) and widely distributed (good diversity). The optimal solution of ENSGA-II is better than the optimal solution of NSGA-II. This shows that ENSGA-II has obvious advantages in small-scale examples. In large-scale examples, as the number of iterations increases, high-quality solutions that are better than NSGA-II can be obtained. Therefore, the above analysis results show that ENSGA-II can obtain a better Pareto optimal solution in a shorter time.

[0091] In summary, this implementation first takes the multi-objective evolutionary algorithm as the theoretical basis. The multi-objective evolutionary algorithm is a mature and perfect technology, and NSGA-II has achieved excellent performance in many fields. The proposed ENSGA-II algorithm, based on the NSGA-II idea, sorts the obtained solutions by Pareto front rank and selects elite solutions based on the knee point criterion to find the approximate optimal solution that performs well in multiple objective dimensions. Compared with the multi-objective evolutionary algorithm based on indicators and decomposition, the ENSGA-II algorithm has the optimal theoretical support of the multi-objective evolutionary algorithm and the knee point criterion, which ensures the search performance and stability of the algorithm. By proposing a new crossover technology to help the search individual conduct extensive exploration in the early stage of iteration, the creative adaptive mutation technology is proposed to enhance the development ability of the search individual, thereby improving the accuracy of the optimal solution. The balance between local search and global search can not only improve the search efficiency of the algorithm, but also improve the search accuracy of the algorithm. This embodiment is based on the elite solution selection criteria of congestion and knee point criterion, which reduces the number of Pareto front solutions in the later stage of iteration, resulting in decision-making difficulties for decision makers, and enhances the practical practicality of the algorithm.

[0092] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of protection of the present disclosure. The attached method claims present the elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.

[0093] In the above detailed description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure should not be interpreted as reflecting an intention that the embodiments of the claimed subject matter require more features than are clearly stated in each claim. On the contrary, as reflected in the appended claims, the invention is in a state of having less than all the features of the disclosed individual embodiments. Therefore, the appended claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.

[0094] Those skilled in the art will also appreciate that the various illustrative logic blocks, modules, circuits, and algorithmic steps described in conjunction with the embodiments herein can all be implemented as electronic hardware, computer software, or a combination thereof. In order to clearly illustrate the interchangeability between hardware and software, various illustrative components, blocks, modules, circuits, and steps are generally described above around their functions. Whether such functions are implemented as hardware or software depends on specific applications and the design constraints imposed on the entire system. A skilled person can implement the described functions in an alternative manner for each specific application, but such implementation decisions should not be interpreted as departing from the scope of protection of the present disclosure.

[0095] The steps of the method or algorithm described in conjunction with the embodiments herein may be directly embodied as hardware, a software module executed by a processor, or a combination thereof. The software module may be located in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a mobile disk, a CD-ROM, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor so that the processor can read information from the storage medium and can write information to the storage medium. Of course, the storage medium may also be an integral part of the processor. The processor and the storage medium may be located in an ASIC. The ASIC may be located in a user terminal. Of course, the processor and the storage medium may also be present in a user terminal as discrete components.

[0096] For software implementation, the techniques described in this application can be implemented with modules (e.g., procedures, functions, etc.) that perform the functions described in this application. These software codes can be stored in a memory unit and executed by a processor. The memory unit can be implemented within the processor or outside the processor. In the latter case, it is coupled to the processor in a communication manner via various means, which are well known in the art.

[0097] The above description includes examples of one or more embodiments. Of course, it is impossible to describe all possible combinations of components or methods for the purpose of describing the above embodiments, but it should be recognized by those skilled in the art that the various embodiments may be further combined and arranged. Therefore, the embodiments described herein are intended to cover all such changes, modifications and variations that fall within the scope of protection of the appended claims. In addition, with respect to the term "comprising" used in the specification or claims, the word is covered in a manner similar to the term "including", just as "including," is explained as a transitional word in the claims. In addition, any term "or" used in the specification of the claims is intended to mean "non-exclusive or".

Claims

1. A method for warehouse replenishment and storage location collaborative planning, characterized in that: include: S100. Construct a warehouse replenishment and storage coordination model, the goal of which is to minimize the total energy consumption of warehouse transportation equipment and the warehouse channel load; the decision variable in the warehouse replenishment and storage coordination model is the variable y of the sub-storage k of shelf l assigned to channel r by SKUp pklr SKUp is assigned to the variable of the sub-storage position i on the channel r1 in the front area SKUp is assigned to the variable of sub-slot j on channel r2 in the pre-stage area Number of SKUp picks in the front-end channel r1 Number of SKUp picks in storage area aisle r2 S200. Solve the warehouse replenishment and storage coordination model based on the improved non-dominated sorting elite genetic algorithm, and obtain the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient based on the knee point criterion.

2. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 1, characterized in that: In S100, the minimum total energy consumption function of the warehouse transportation equipment is: Where R represents the channel set, r=1,…|R|, r∈R; r1∈R1 represents the channel index of the front area, r2∈R2 represents the channel index of the storage area; I represents the sub-storage location set of each channel in the front area, i=1,…|I|, i∈I; J represents the sub-storage location set of each channel in the storage area, j=1,…|J|, j∈J; P represents the SKU set, p=1,…|P|, p∈P; It represents the energy consumption required by the AGVs to transport SKUp from storage location i in the front area r1 to the picking platform. Indicates the picking frequency of SKUp in each sub-storage location in the front area. Indicates that SKUp is allocated to the sub-storage position i on the channel r1 in the front area; represents the energy consumption required by the robot to transport SKUp from storage location j in storage area r2 to the picking station in the front area; Indicates the picking frequency of SKUp in each sub-storage location in the storage area. Indicates that SKUp is allocated to sub-storage location j on storage channel r2; represents the energy consumption required by the robot to move SKUp from storage location j in storage area r2 to storage location i in the front area r1; It indicates the average replenishment frequency of each sub-storage location allocated by SKUp from the picking platform to the front area. It indicates the average replenishment frequency of SKUp from each sub-location allocated in the storage area to the picking station.

3. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 2, characterized in that: In S100, the warehouse channel load minimum function is: in, is the number of times SKUp is picked in the front-end channel r1, is the average number of picking times in all channels of the front area, is the number of times the SKU is picked in the storage area channel r2, is the average number of picks across all aisles in the storage area.

4. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 3, characterized in that: In S100, the constraints of the minimum total energy consumption function of the warehouse transportation equipment and the minimum warehouse channel load function include at least: one sub-storage location can only store one SKU, the number of SKU types stored on a single shelf cannot exceed the total sub-storage locations of the single shelf, and the number of sub-storage locations occupied by each SKU cannot exceed the total allocated storage locations.

5. The method for warehouse replenishment and storage space collaborative planning as claimed in claim 1, characterized in that: In S200, the warehouse replenishment and storage location coordination model is solved based on an improved non-dominated sorting elite genetic algorithm to obtain an optimal population solution and a corresponding minimum energy consumption value and channel usage difference coefficient. The specific method includes: S201. Set parameters and initialize the population; S202. Generating a selected population based on the initialization population; S203. Generate a crossover population based on the crossover method; S204. Correcting the crossover population using a correction method for an infeasible solution to obtain a corrected population; S205. mutating the corrected population using a mutation method to obtain a mutated population; S206. Merge the initial population and the mutated population to form a merged population, perform non-dominated sorting on the merged population, and obtain the Pareto rank of each solution in the merged population; S207. Based on the Pareto level of each solution in the merged population, a new offspring is generated, the number of iterations is updated, and it is determined whether the current number of iterations reaches the maximum number of iterations. If the maximum number of iterations is reached, the process goes to S208; if the maximum number of iterations is not reached, the process returns to S202; S208. The current number of iterations reaches the maximum number of iterations, and the optimal population solution and the corresponding minimum energy consumption value and channel usage difference coefficient are obtained.

6. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 5, characterized in that: In S201, the parameters are set and the population is initialized. The specific method includes: setting the parameters, the parameters at least including the total number of storage spaces in the front area and the storage area, the number of empty storage spaces and the number of non-empty storage spaces, and the number of sub-storage spaces allocated to each SKU; determining the initial solution population size according to the parameters, randomly selecting a SKU and placing it in the sub-storage spaces in the front area and the storage area until the number of sub-storage spaces allocated to the SKU is allocated, and then allocating storage spaces to the next SKU, and repeating the above steps until storage space allocation is completed for all SKUs.

7. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 5, characterized in that: In S202, a selection population is generated based on the initialization population. The specific method includes: performing non-dominated sorting on the initialization population to obtain a Pareto front set; judging the quality of the solution according to the quality of the Pareto front set, and using a binary tournament technique to select elite individuals until the elite individuals reach a preset solution population size. At this time, the population composed of elite individuals is the selection population.

8. The method for warehouse replenishment and storage space collaborative planning as claimed in claim 5, characterized in that: In S203, a crossover population is generated based on the crossover method, and the crossover method includes a crossover method based on an optimal target and a two-point crossover method; wherein the crossover method based on an optimal target specifically includes: randomly selecting the optimal chromosome S on two target dimensions based on the selected population; best1 and S best2 , respectively select the part between the preamble area and the storage area and the corresponding part of a parent chromosome Parent1 for crossover to generate a new offspring offspring1; the two-point crossover method specifically includes: randomly selecting genes between two parent chromosomes for exchange to obtain two new offspring chromosomes offspring2 and offspring3; until all traversals complete all selected population solutions, that is, a crossover population is generated.

9. A method for warehouse replenishment and storage space collaborative planning as claimed in claim 5, characterized in that: In S204, the crossover population is corrected by using a correction method for an infeasible solution to obtain a corrected population. The specific method includes: checking the number of sub-storage positions occupied by each SKU in the chromosome in turn, eliminating redundant SKUs, and replacing SKUs with missing sub-storage positions until the number of storage positions occupied by all SKUs meets the allocated number of sub-storage positions, at which point the model has a feasible solution.

10. The method for warehouse replenishment and storage space collaborative planning according to claim 5, characterized in that: In S205, a mutation method is used to mutate the corrected population to obtain a mutated population. The specific method includes: selecting a parent chromosome in the crossover population according to a preset mutation probability for a mutation operation, and randomly selecting genes in the pre-region and the storage region for exchange to complete the mutation operation, until all crossover population solutions are traversed and completed, that is, a mutant population is generated.

Images