Bi-GA-based multi-center warehouse joint replenishment and transportation decision-making method and service platform

Abstract

The invention relates to a Bi-GA-based multi-center warehouse joint replenishment and transportation decision-making method and service platform, and the method comprises the steps: determining the current demands of a plurality of center warehouses for a plurality of target products in a plurality of execution cycles of a decision-making stage after the current; the central warehouse, the execution cycle, the target product and the current demand quantity are coded and initialized, a plurality of multi-dimensional coding bodies are generated, and the multi-dimensional coding bodies comprise a plurality of replenishment decision sub-coding bodies; the replenishment decision coding sub-body comprises a plurality of replenishment decision coding sub-bodies which are used for representing replenishment decisions of the corresponding center bins for replenishment of various target products in each execution period; and performing replenishment path planning decision iteration of an inner layer and replenishment decision iteration of an outer layer on the basis of the replenishment decision sub-coding body corresponding to the multi-dimensional coding body by using a preset double-layer genetic algorithm until a target multi-dimensional coding body is generated, and taking all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional coding body as decision results.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent logistics and warehouse management, in particular to a multi-center warehouse joint replenishment and transportation decision-making method and service platform based on Bi-GA. BACKGROUND

[0002] The e-commerce self-operation mode refers to that an e-commerce enterprise realizes closed-loop whole-process management of each link of commodity selection and purchase, transportation, warehousing and sales through independent decision-making; in order to improve the distribution efficiency, the self-operation e-commerce usually adopts a multi-level distributed warehousing mode to realize rapid response to demand within a regional range; in the related technology, the self-operation e-commerce constructs a purchase-transportation-warehousing system composed of multiple central warehouses, multiple suppliers, multiple products and transportation fleets, and the core of the purchase-transportation-warehousing decision of the self-operation e-commerce lies in replenishment planning; the existing joint replenishment problem (JRP) is usually based on deterministic demand and fixed replenishment rules, and the fixed replenishment rules include fixed period or fixed quantity, but the replenishment with fixed period or fixed quantity cannot effectively cope with market fluctuations; at the same time, the replenishment quantity is closely coupled with the transportation path of the transportation vehicle in the replenishment planning decision of the self-operation e-commerce, the existing replenishment planning adopts the split of replenishment decision and path planning, which easily leads to inventory imbalance, inventory explosion and high total system cost.

[0003] At present, there is no effective solution to the problem of weak ability to cope with dynamic interference, easy to lead to inventory imbalance and explosion in the replenishment decision of the purchase-transportation-warehousing of the self-operation e-commerce in the related technology. SUMMARY

[0004] The embodiments of the present application provide a multi-center warehouse joint replenishment and transportation decision-making method and service platform based on Bi-GA, to at least solve the problem of weak ability to cope with dynamic interference, easy to lead to inventory imbalance and explosion in the replenishment decision of the purchase-transportation-warehousing of the self-operation e-commerce in the related technology.

[0005] In a first aspect, the embodiments of the present application provide a multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA, comprising: determining current demand quantities of a plurality of target products to be replenished in a plurality of execution cycles corresponding to a decision stage after the current in a plurality of center warehouses, wherein the current demand quantity is a predicted quantity of goods sales obtained according to current inventory quantities of the plurality of center warehouses by using a preset sales prediction model, the sales prediction model is a model based on PSO-ALSTM training, and the plurality of execution cycles corresponding to a decision stage after the current are generated by rolling a plurality of execution cycles corresponding to a last decision stage by one execution cycle; encoding and initializing the center warehouses, the execution cycles, the target products and the current demand quantities according to a preset encoding manner to generate a plurality of multi-dimensional encoding bodies, wherein the multi-dimensional encoding bodies include a plurality of replenishment decision sub-encoding bodies, one replenishment decision sub-encoding body is associated with one center warehouse, the replenishment decision sub-encoding body includes a plurality of replenishment decision encoding sub-bodies for representing replenishment decisions of the corresponding center warehouse in each execution cycle for a plurality of target products, and the replenishment decision at least includes a replenishment quantity of the target product; using a preset double-layer genetic algorithm to perform inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration based on the replenishment decision sub-encoding bodies corresponding to the plurality of multi-dimensional encoding bodies until a target multi-dimensional encoding body is generated, and using all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional encoding body as decision results, wherein the target replenishment path is generated by using an inner-layer genetic algorithm of the double-layer genetic algorithm to perform replenishment path planning on the replenishment decision encoding sub-bodies of the replenishment decision sub-encoding bodies of the target multi-dimensional encoding body.

[0006] In a second aspect, the embodiments of the present application provide a service platform, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA when executing the computer program.

[0007] Compared with the related art, the multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA provided by the embodiment of the application adopts determining the current demand quantity of a plurality of target products to be replenished in a plurality of execution cycles corresponding to a plurality of center warehouses in a decision-making stage after the current; the center warehouse, the execution cycle, the target product, and the current demand quantity are encoded and initialized according to a preset encoding mode to generate a plurality of multi-dimensional encoding bodies, the multi-dimensional encoding body includes a plurality of replenishment decision sub-encoding bodies, one replenishment decision sub-encoding body is associated with one center warehouse, the replenishment decision sub-encoding body includes a plurality of replenishment decision encoding sub-bodies for representing the replenishment decision of the corresponding center warehouse in each execution cycle for a plurality of target products, and the replenishment decision at least includes the replenishment quantity of the target product; a preset double-layer genetic algorithm is used to perform inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration based on the replenishment decision sub-encoding bodies corresponding to the plurality of multi-dimensional encoding bodies, until a target multi-dimensional encoding body is generated, and all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional encoding body are used as the decision result, the target replenishment path is generated by using the inner-layer genetic algorithm of the double-layer genetic algorithm to perform replenishment path planning on the replenishment decision encoding sub-bodies of the replenishment decision sub-encoding bodies of the target multi-dimensional encoding body; the prediction, replenishment, and transportation decisions are fused and cooperatively optimized to construct a running mechanism of prediction, decision, execution, and update, and combined with a flexible replenishment strategy and a heterogeneous vehicle fleet path to jointly optimize replenishment, at the same time, the complexity of the planning decision result solution and the expansion of the prediction error are reduced through the double-layer genetic algorithm and the rolling horizon method, the replenishment efficiency and response capability of the mining, transportation, and storage system are improved, demand fluctuations are adapted, the replenishment and transportation path decision effect is good, the operation cost is reduced, and the problems of weak dynamic interference response capability, easy inventory imbalance, and warehouse explosion in the replenishment decision of the mining, transportation, and storage of the self-operated e-commerce in the related art are solved.

[0008] The details of one or more embodiments of the application are presented in the following drawings and description to make other features, objects, and advantages of the application more apparent. BRIEF DESCRIPTION OF DRAWINGS

[0009] The drawings described herein are intended to provide further understanding of the application, and form a part of the application. The illustrative embodiments of the application and their description serve to explain the application without constituting an improper limitation of the application. In the drawings: Figure 1 is a hardware structure block diagram of a terminal of the multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA of the embodiment of the application; Figure 2 is a flowchart of the multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA according to the embodiment of the application; Figure 3 is a fixed mapping relationship matrix diagram between the target product and the target supply object of the embodiment of the present application; Figure 4 is a schematic diagram of a multi-dimensional encoding body of the embodiment of the present application; Figure 5 is a schematic diagram of a replenishment decision sub-encoding body of the embodiment of the present application; Figure 6 is a schematic diagram of path segmentation of a node sequence of the embodiment of the present application; Figure 7 is a schematic diagram of an access encoding sub and a cross range of the embodiment of the present application; Figure 8 is a schematic diagram of an access encoding sub and an empty encoding sub of the embodiment of the present application; Figure 9 is a schematic diagram of an access encoding sub performing cross operation of the embodiment of the present application; Figure 10 is a schematic diagram of two first target sub-encoding bodies of the embodiment of the present application; Figure 11 is a schematic diagram of a first replenishment decision sub-encoding body generated after product-level cross operation of the embodiment of the present application; Figure 12 is another schematic diagram of two second target sub-encoding bodies of the embodiment of the present application; Figure 13 is a schematic diagram of a second replenishment decision sub-encoding body generated after period-level cross operation of the embodiment of the present application; Figure 14 is a structural block diagram of the self-operated e-commerce multi-center warehouse joint replenishment and transportation decision device based on cloud-edge collaboration of the embodiment of the present application. DETAILED DESCRIPTION

[0010] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application is described and explained below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present application. In addition, it can be understood that although the efforts made in this development process can be complex and lengthy, some design, manufacture or production changes made on the basis of the technical content disclosed in the present application are only routine technical means for those of ordinary skill in the art related to the content disclosed in the present application, and should not be understood as insufficient disclosure of the content disclosed in the present application.

[0011] Reference to an "embodiment" in this application means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. It is expressly understood that that the embodiments described herein are merely examples from among a great variety of embodiments that, as a rule, are combinable with one another.

[0012] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the ordinary meanings as understood by one of ordinary skill in the art to which this application pertains. The terms "a", "an", "one", "this", and similar referents in the context of describing the application are to be construed to be inclusive, both singular and plural, unless otherwise indicated. The terms "comprising", "containing", "having", and "including" and any variations thereof used herein are intended to cover a non-exclusive inclusion such that a process, method, system, product, or apparatus that comprises, contains, has, or includes a list of steps or modules (units) is not limited to those steps or modules which are recited, but can also include additional steps or modules that are not expressly listed or can also include additional steps or modules that are inherent in such process, method, product, or apparatus. The term "multiple stages" refers to more than or equal to two stages. The term "and / or" describes an association relationship of associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that A exists alone, A and B exist together, and B exists alone. The terms "first", "second", "third", and the like are merely used to distinguish similar objects, and do not represent a specific order for the objects.

[0013] Before the embodiments of the present application are described, the following definitions of related symbols and parameters involved in the embodiments of the present application are defined as follows: p represents the pth target product, p∈P; P represents a target product set, P={1, 2, 3, …, |P|}; s represents the sth target supply object, s∈S; S represents a target supply object set, S={1, 2, 3, …, |S|}; w represents the wth central warehouse, w∈W; W represents a central warehouse set, W={1, 2, 3, …, |W|}; r, s represent vehicle access node indexes, r, s∈{w}∪S; k represents the kth replenishment vehicle, k∈K; K represents a vehicle set; t represents the tth execution cycle, t∈T; T represents an execution cycle set, T={1, 2, 3, …, |T|}; T1 represents a decision cycle set for replenishment quantity and replenishment path decision, T1={1, 2, 3, …, T-1}; T2 represents a calculation cycle set for sales behavior, T2={1, 2, 3, …, T}; B ps represents a correspondence between a target product and a target supply object, p∈P, s∈S, if the target product p is provided by the target supply object s, B ps= 1, otherwise, B ps = 0; denotes the predicted demand of the target product p in the wth central warehouse in the tth execution cycle, t e T2, w e W, p e P; denotes the purchase price of the target product p with price discount, p e P, t e T1, ; sp p denotes the sales price of the target product p; v p denotes the volume of the target product p, p e P; sc w denotes the storage cost per unit volume of product in the wth central warehouse per execution cycle; denotes the maximum capacity of the wth central warehouse; lc p denotes the out-of-stock loss penalty coefficient of the target product p in the central warehouse; d rs denotes the distance between the vehicle access node r and the vehicle access node s, r, s e {w} U S, r ≠ s; fc k denotes the fixed departure cost of the replenishment vehicle k; vc k denotes the unit distance travel cost of the replenishment vehicle k; denotes the maximum capacity of the replenishment vehicle k; denotes the replenishment amount of the target product p to the wth central warehouse in the tth execution cycle, p e P, w e W, t e T1; denotes whether the kth vehicle departs from the wth central warehouse, travels along the path r, s in the tth execution cycle, r, s e {w} U S, r ≠ s, w e W, k e K, if yes, , otherwise, ; denotes whether the central warehouse w is enabled with the vehicle k in the tth execution cycle, if yes, , otherwise, , w e W, t e T2, k e K; denotes whether the vehicle departing from the central warehouse w in the tth execution cycle visits the sth target supply object, if yes, , otherwise, , w e W, s e S, t e T1; denotes the remaining inventory of the target product p in the central warehouse w at the end of the tth execution cycle, w e W, p e P, t e T1 n T2; denotes the total volume of the product replenished at the target supply object s in the central warehouse w in the tth execution cycle, w e W, s e S, t e T1; denotes the total volume of the product replenished at the target supply object s by the replenishment vehicle K of the central warehouse w in the tth execution cycle, w e W, s e S, t e T1

[0014] The embodiments of the present application are described as follows: The method provided in the embodiment can be executed in a terminal, a computer or a similar computing device. Taking a terminal as an example, Figure 1 is a hardware structure block diagram of a terminal of the Bi-GA-based multi-center warehouse joint replenishment and transportation decision method of the embodiment of the application. As shown in Figure 1 , the terminal can include one or more (only one is shown in Figure 1 ) processors 102 (the processor 102 can include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the terminal can further include a transmission device 106 for communication function and an input and output device 108. Those skilled in the art can understand that Figure 1 the structure shown is only schematic, which does not limit the structure of the terminal. For example, the terminal can further include more or fewer components than those shown in Figure 1 , or have a different configuration from that shown in Figure 1 .

[0015] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the Bi-GA-based multi-center warehouse joint replenishment and transportation decision method of the embodiment of the application. The processor 102 executes various function applications and data processing by running the computer program stored in the memory 104, that is, implements the method described above. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the terminal 10 through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0016] The transmission device 106 is used to receive or send data via a network. Specific examples of the network can include a wireless network provided by a communication provider of the terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC) which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, RF) module which is used to communicate with the Internet in a wireless manner.

[0017] The embodiment provides a Bi-GA-based multi-center warehouse joint replenishment and transportation decision method running on the terminal, Figure 2is a flowchart of a multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA according to an embodiment of the present application, as shown in Figure 2 The flowchart includes the following steps: In step S201, current demand quantities of target products to be replenished in a plurality of execution periods corresponding to a next decision stage after the current decision stage are determined for a plurality of center warehouses, wherein the current demand quantity is a predicted product sales quantity obtained according to a current inventory quantity of the plurality of center warehouses by using a preset sales prediction model, the sales prediction model is a model based on PSO-ALSTM, and the plurality of execution periods corresponding to the next decision stage are generated by rolling one execution period for the plurality of execution periods corresponding to the last decision stage.

[0018] In the present embodiment, the execution subject of the decision-making method of the present application includes but is not limited to a control terminal, a control system or a cloud platform controlled by a self-operated e-commerce enterprise for the plurality of center warehouses. In the present embodiment, after the execution subject perceives the replenishment and transportation decisions of the plurality of center warehouses in response to the first execution period of the last decision stage, the execution subject will start to execute the replenishment and transportation decisions of the plurality of execution periods of the next decision stage, that is, the replenishment and transportation decision-making method of the present application adopts a method of planning the replenishment and transportation decisions of the plurality of execution periods again after one execution period is executed. At the same time, when the planning is performed again, the demand quantities of the center warehouses that need to be replenished from the target supply objects in the plurality of execution periods will be predicted again. However, in order to reduce the prediction error and avoid the amplification of the error, the input data will be the saleable inventory quantities of the center warehouses after the replenishment and transportation decisions of the first execution period of the last decision stage are responded to, and the output product sales quantity will be the current demand quantity of the corresponding center warehouse. It can be understood that the actual demand quantity or the out-of-warehouse quantity of the corresponding center warehouse in the first execution period of the last decision stage can be determined according to the saleable inventory quantity of the center warehouse and the product demand quantity of the center warehouse decided in the first execution period of the last decision stage, and the product sales quantity as the current demand quantity can be predicted by inputting the corresponding out-of-warehouse quantity into the sales prediction model. After the corresponding current demand quantity is obtained, the replenishment and transportation decisions of the plurality of center warehouses in the plurality of execution periods will be performed according to the corresponding current demand quantity.

[0019] In the embodiment, the allocation quantity of each target product of each node warehouse (warehouse facing the consumer terminal object) in each execution cycle is calculated, i.e., the outbound quantity of each target product of the central warehouse in each execution cycle (to-be-executed execution cycle) is calculated, so as to realize the conversion of the demand prediction results of each region (node warehouse) into the replenishment quantity of the central warehouse; in the embodiment, the sales prediction model based on PSO-ALSTM is used to predict the sales of the goods of the node warehouse or the store in front of the node warehouse in each execution cycle in the future, and the demand quantity of the replenishment quantity of the central warehouse is converted into the replenishment quantity of the central warehouse through the EOQ model; it should be understood that in the final results of the procurement decision and the transportation decision, the final replenishment quantity of the target product can be the same as the replenishment quantity corresponding to the set demand information, but can also be different, but will not exceed the total demand quantity.

[0020] In the embodiment, the sales prediction model uses the particle swarm optimization (PSO) method to construct an adaptive structure LSTM model, and uses the historical sales data of each node warehouse and various feature parameters (for example: product seasonal attribute parameters, product transaction attribute parameters) as the training data set and the test data set for training, the sales prediction model uses the deep learning framework TensorFlow, and uses the high-level interface Keras to construct, the sales prediction model is trained to output the demand quantity of the target product of the node warehouse in the future corresponding time period according to the input sales data and feature parameters of the target product.

[0021] In step S202, the central warehouse, the execution cycle, the target product and the current demand quantity are encoded and initialized according to a preset encoding mode to generate a plurality of multi-dimensional encoding bodies, wherein the multi-dimensional encoding body includes a plurality of replenishment decision sub-encoding bodies, one replenishment decision sub-encoding body is associated with one central warehouse, the replenishment decision sub-encoding body includes a plurality of replenishment decision encoding sub-bodies for representing the replenishment decision of the corresponding central warehouse to replenish a plurality of target products in each execution cycle, and the replenishment decision at least includes the replenishment quantity of the target product.

[0022] In the embodiment, after determining the current demand quantity corresponding to each central warehouse, a corresponding encoding is performed according to the current demand quantity and population initialization is performed, a plurality of multi-dimensional encoding bodies are generated, so as to realize the replenishment path planning of the inner layer and the replenishment decision of the outer layer by using the double-layer genetic algorithm; in the embodiment, one replenishment decision sub-encoding body is a replenishment decision scheme corresponding to one central warehouse, one replenishment decision sub-encoding body includes a plurality of replenishment decision encodings, one replenishment decision encoding is a decision of replenishing a plurality of target products in one execution cycle, and one replenishment decision encoding includes a plurality of decision sub-codes including product encodings and replenishment quantity encodings, the replenishment quantity encoding is used to represent the replenishment quantity of one target product in one execution cycle, and when encoding, the value of the replenishment quantity encoding is determined based on the value interval determined based on the determined current demand quantity of the target product; in the embodiment, one product encoding is associated with one target supply object; in the embodiment, the multi-dimensional encoding body is organized in a matrix form as a three-dimensional structure of central warehouses (W) x target products (P) x execution cycles (T), one central warehouse encoding is in a two-dimensional structure of target products (P) x execution cycles (T), and the gene value with the replenishment quantity encoding as the core is a non-negative integer, which represents the replenishment quantity of the target product by the specific central warehouse in the specific execution cycle (for example, the first execution cycle), and before encoding the product encoding, a fixed mapping relationship matrix between the target product and the target supply object is established, as shown in Figure 3 , when the replenishment quantity of any target product provided to a certain target supply object in a certain execution cycle is greater than zero, the access demand to the target supply object in the execution cycle is triggered, so as to organically connect the replenishment decision and the path decision, and the initial plurality of multi-dimensional encoding bodies are randomly generated under the premise of satisfying the vehicle capacity constraint, for example, taking 3 central warehouses, 5 target products, 2 target supply objects (the corresponding mapping relationship is shown in Figure 3 ), and 7 execution cycles as an example, one multi-dimensional encoding body in the encoding is shown in Figure 4 , wherein one replenishment decision sub-encoding body is shown in Figure 5 ; in the embodiment, Figure 4 and Figure 5 , the target product column and the replenishment quantity encoding column in one execution cycle form a replenishment decision encoding, it should be noted that when encoding, only one column of target products is presented, and when decoding or processing, the replenishment decision encoding formed by one column of target products and one column of picking quantity data is presented; it should be understood that each target product is associated with a corresponding target supply object, that is, after the number of the target product is determined, the target supplier corresponding to the target product is determined (as shown in Figure 3 ).

[0023] In step S203, the preset double-layer genetic algorithm is used to perform inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration based on the replenishment decision sub-codes of the multiple multi-dimensional codes, until a target multi-dimensional code is generated, and all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional code are taken as the decision result, wherein the target replenishment path is generated by using the inner-layer genetic algorithm of the double-layer genetic algorithm to perform replenishment path planning on the replenishment decision code of the replenishment decision sub-code of the target multi-dimensional code.

[0024] In this embodiment, the outer-layer genetic algorithm of the double-layer genetic algorithm is responsible for determining the replenishment quantity of each target product, each central warehouse and each execution cycle, adopts real number coding mode, each corresponding code represents a complete replenishment scheme for multiple central warehouses, and each replenishment decision sub-code represents a complete replenishment scheme for one central warehouse; in this embodiment, for the replenishment decision of the outer layer, the tournament selection method is used for parent selection, the two-point crossover operation is used for recombination of replenishment decisions, and the picking quantity coding variation based on vehicle capacity constraint is used to maintain population diversity, and the fitness evaluation of the genetic evolution operation of the outer layer comprehensively considers the replenishment cost and the transportation cost from the inner layer; in this embodiment, the fitness function is to maximize the benefit, and the parameters affecting the fitness include the sales revenue generated by the demand quantity of the target product after the sale of the target product, the cost of purchasing the demand quantity of the target product, the central warehouse storage cost, and the shortage cost caused by the shortage of the target product, for multiple parameters, the preset fitness function is used for calculation, and then the fitness corresponding to one multi-dimensional code is determined; in this embodiment, the inner-layer GA of the double-layer genetic algorithm is a replenishment path planning decision layer: solving the capacitated vehicle routing problem (CVRP) for the given replenishment plan of the outer layer, which adopts integer coding, the codes of the target supply objects associated with the target products of each replenishment decision code are arranged into corresponding path node sequences, and the path planning is performed according to the reference Figure 6 , it is assumed that in the current replenishment scheme, 5 target supply objects need to be visited in one execution cycle, and all the nodes are arranged into integers, and after the genetic operation, the node codes are allocated vehicles according to the constraint conditions, and Figure 6 It can be known that a total of 2 replenishment vehicles are allocated, and thus two sub-paths satisfying the vehicle maximum capacity constraint condition are formed; the replenishment codes corresponding to one multi-dimensional code also apply the tournament selection, two-point crossover and bit mutation to optimize the transportation cost under the capacity and path constraints; in this embodiment, a feedback mechanism is also adopted, that is, the inner-layer GA returns the optimized transportation cost to the outer layer, and the outer layer updates the overall fitness evaluation accordingly, and the iteration process continues until the termination condition is met, so as to ensure that the replenishment and path decisions are jointly optimized rather than processed independently.

[0025] By the above steps S201 to S203, the current demand of the target product to be replenished is predicted in a plurality of execution cycles corresponding to a decision stage after the current in the plurality of central warehouses, wherein the current demand is the predicted sales volume of the goods according to the current inventory of the plurality of central warehouses by using a preset sales volume prediction model, the sales volume prediction model is a model based on PSO-ALSTM, and the plurality of execution cycles corresponding to a decision stage after the current is generated by rolling the plurality of execution cycles corresponding to the last decision stage by one execution cycle; the central warehouse, the execution cycle, the target product and the current demand are encoded and initialized according to a preset encoding method to generate a plurality of multi-dimensional encoding bodies, wherein the multi-dimensional encoding body includes a plurality of replenishment decision sub-coding bodies, one replenishment decision sub-coding body is associated with one central warehouse, the replenishment decision sub-coding body includes a plurality of replenishment decision coding sub-codes for representing the replenishment decision of the corresponding central warehouse in each execution cycle for the plurality of target products, and the replenishment decision at least includes the replenishment quantity of the target product; by using a preset double-layer genetic algorithm, based on the replenishment decision sub-coding bodies corresponding to the plurality of multi-dimensional encoding bodies, the inner-layer replenishment path planning decision iteration and the outer-layer replenishment decision iteration are performed until the target multi-dimensional encoding body is generated, and all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional encoding body are taken as the decision result, wherein the target replenishment path is generated by using the inner-layer genetic algorithm of the double-layer genetic algorithm to perform replenishment path planning on the replenishment decision coding sub-codes of the replenishment decision sub-coding bodies of the target multi-dimensional encoding body; the prediction, replenishment and transportation decisions are fused and cooperatively optimized to construct a running mechanism of prediction, decision, execution and update, combined with flexible replenishment strategy and heterogeneous vehicle fleet path to jointly optimize the replenishment, at the same time, by using the double-layer genetic algorithm and the rolling horizon method, the complexity of the planning decision result solving is reduced and the expansion of the prediction error is inhibited, the replenishment efficiency and response ability of the mining, transportation and storage system are improved, the demand fluctuation is adapted, the replenishment and transportation path decision effect is good, the operation cost is reduced, and the problems of weak dynamic interference response ability, easy to cause inventory imbalance and out-of-stock in the replenishment decision of the mining, transportation and storage of the self-operated e-commerce in the related technology are solved.

[0026] It should be noted that the embodiments of the present application construct an integrated solution of "prediction-decision-execution-update" to balance the prediction accuracy, calculation efficiency and system response ability; by designing a double-layer genetic algorithm and a rolling horizon method, the calculation complexity is managed and the demand fluctuation is adapted.

[0027] In some embodiments, by using a preset double-layer genetic algorithm, based on the replenishment decision sub-coding bodies corresponding to the plurality of multi-dimensional encoding bodies, the inner-layer replenishment path planning decision iteration and the outer-layer replenishment decision iteration are performed, including the following steps: Step 21, using the inner genetic algorithm of the double-layer genetic algorithm, the all replenishment path planning of the replenishment decision coding sub corresponding to the multiple multi-dimensional coding bodies participating in the replenishment decision iteration of this time is carried out, and the target transportation path corresponding to each replenishment decision coding sub is obtained.

[0028] In this embodiment, the inner genetic algorithm of the double-layer genetic algorithm carries out replenishment path decision for the replenishment plan given by the outer layer (corresponding to all replenishment decision coding subs). After receiving the replenishment plan transmitted by the outer layer, the inner genetic algorithm converts it into the target supply object access demand of each execution cycle, and then solves the optimal replenishment path under this replenishment plan. Considering the independence of the transportation resources of each central warehouse, the parallel optimization strategy is used to process the path decision of different central warehouses. The access sub coding body and the coding of the access coding for replenishment path decision are integer sequences, which correspond to the access order of node coding. Each node coding in the node sequence corresponds to a target supply object. The complete sub-transportation path is formed by the central warehouse starting to access each node coding in turn and returning to the central warehouse. In this embodiment, the coding of the target supply objects associated with the target products of each replenishment decision coding sub is arranged into the corresponding node sequence, and the coding of the target supply objects associated with the target products of each replenishment decision coding sub is arranged into the corresponding node sequence. Referring to Figure 6 , it is assumed that in the current replenishment scheme, 5 target supply objects need to be accessed in a certain execution cycle, all of which are nodes, and all node codings are arranged into integers. After genetic operation, according to the constraint condition, the node coding is allocated vehicles, and Figure 6 It can be seen that a total of 2 replenishment vehicles are allocated, so two sub-paths (0-3-2-0 and 0-5-4-1-0) that satisfy the vehicle maximum capacity constraint condition are formed.

[0029] Step 22, after determining the fitness of each multi-dimensional coding body according to the sub-fitness corresponding to each replenishment decision sub-coding body, using the outer genetic algorithm of the double-layer genetic algorithm, at least one set of outer genetic evolution operations is carried out on the multiple multi-dimensional coding bodies participating in this iteration, to generate multiple multi-dimensional coding bodies completing this replenishment decision iteration, wherein the sub-fitness is determined according to the replenishment transportation cost of all target transportation paths corresponding to the replenishment decision sub-coding body and the income parameter corresponding to the replenishment decision sub-coding body.

[0030] In the present embodiment, the replenishment decision of the outer layer uses the tournament selection method for parent selection, adopts two-point crossover operation to recombine the replenishment decision, and maintains population diversity through the picking quantity coding mutation based on the vehicle capacity constraint. The fitness evaluation of the genetic evolution operation of the outer layer comprehensively considers the replenishment cost and the transportation cost from the inner layer. The fitness of the outer layer is maximization of the profit. The parameters affecting the fitness include the sales revenue generated by the sales of the demand quantity of the target product in the central warehouse, the cost of purchasing the demand quantity of the target product, the storage cost of the central warehouse, and the shortage cost caused by the shortage of the target product. For multiple parameters, a preset fitness function is used to calculate the fitness corresponding to a multi-dimensional coding body.

[0031] Step 23, based on the multiple multi-dimensional coding bodies completing the replenishment decision iteration, the steps of planning the corresponding target transportation path using the corresponding inner layer genetic algorithm, determining the fitness corresponding to the corresponding multi-dimensional coding body, and performing the set outer layer genetic evolution operation using the corresponding outer layer genetic algorithm are repeatedly executed until multiple multi-dimensional coding bodies with a fitness greater than a fitness threshold are generated.

[0032] Step 24, among the multiple multi-dimensional coding bodies with a fitness greater than the fitness threshold, the multi-dimensional coding body with the maximum fitness is selected to obtain a target multi-dimensional coding body.

[0033] Through the above steps 21 to 24, the inner layer replenishment path planning decision iteration and the outer layer replenishment decision iteration are realized according to the coded multiple multi-dimensional coding bodies using the double-layer genetic algorithm In some embodiments, the inner layer genetic algorithm of the double-layer genetic algorithm performs replenishment path planning on all replenishment decision coding sub-bodies corresponding to the multiple multi-dimensional coding bodies participating in the replenishment decision iteration, which is realized through the following steps: Step 31, obtaining the replenishment decision sub-coding body corresponding to each multi-dimensional coding body participating in the replenishment decision iteration, and determining all replenishment decision coding sub-bodies of each replenishment decision sub-coding body, wherein the replenishment decision coding sub-body includes multiple product coding and replenishment quantity coding sub-decisions, the replenishment quantity coding is used to represent the replenishment quantity of a target product in an execution cycle, and one product coding is associated with one target supply object.

[0034] Step 32, encode all the target supply objects associated with the product codes of each replenishment decision subcode as node codes, encode the node sequence, encode one node sequence as an access subcode, encode the access subcode corresponding to the replenishment decision subcode, and generate a replenishment code corresponding to the access subcode, wherein one replenishment code corresponds to one multi-dimensional code, based on the vehicle capacity constraint, the node sequence corresponding to the access subcode is path segmented to generate at least one sub-transportation path, and the candidate transportation path corresponding to the access subcode includes at least one sub-transportation path, and the replenishment quantity of each node code corresponding to the node replenishment is determined according to the replenishment quantity code corresponding to all product codes associated with a target supply object.

[0035] In this embodiment, the access subcode of the replenishment path decision and the access code are encoded by using an integer sequence, which corresponds to the access order of the node code. Each node code in the node sequence corresponds to a target supply object. The complete sub-transportation path starts from the central warehouse, visits each node code in turn, and returns to the central warehouse. In this embodiment, the codes of the target supply objects associated with the target products of each replenishment decision subcode are arranged into a corresponding node sequence, and the reference Figure 6 , assuming that in the current replenishment scheme, 5 target supply objects need to be visited as corresponding nodes in a certain execution period, and all node codes are arranged in integers, after genetic operation, the node codes are allocated vehicles according to the constraint conditions, and Figure 6 It can be seen that a total of 2 replenishment vehicles are allocated, and thus two sub-paths (0-3-2-0 and 0-5-4-1-0) satisfying the vehicle maximum capacity constraint condition are formed.

[0036] Step 33, after determining the replenishment cost parameter corresponding to the candidate transportation path according to the sub-transportation path corresponding sub-replenishment cost parameter, determining the decision fitness corresponding to the replenishment code according to the sub-body replenishment cost parameter of the access subcode determined based on the replenishment cost parameter, and based on the decision fitness, using the inner genetic algorithm to perform a set of inner genetic evolution operations on the access subcodes of the plurality of replenishment codes, until at least one candidate replenishment code with a decision fitness less than a set fitness value is generated, wherein the inner genetic evolution operation includes the following operations: tournament selection operation, two-point crossover operation and exchange mutation operation.

[0037] In this embodiment, the fitness function calculation method of the inner genetic algorithm is to calculate the transportation cost under the corresponding sub-transportation path, including the fixed transportation cost and the variable transportation cost.

[0038] In some optional embodiments, the set of inner genetic evolution operations on the access subcodes of the plurality of replenishment codes in step 33 includes the following steps: Step 331, after selecting two replenishment code bodies from the plurality of replenishment code bodies as parent code bodies, selecting two access child code bodies at the same ordering position from all access child code bodies of the two parent code bodies, and randomly selecting at least one access code at the same ordering position from all access code children of the two corresponding access child code bodies, obtaining at least one parent code child group, and generating a corresponding empty code child group corresponding to the corresponding parent code child group.

[0039] Step 332, randomly selecting a preset number of node codes at the same first coding position from all node codes corresponding to the first parent code child and the second parent code child of the parent code child group, obtaining a first node code set and a second node code set.

[0040] Step 333, filling all node codes of the first node code set to the first coding position corresponding to the first empty code child of the empty code child group, and sequentially filling the node codes of the second parent code child except the first node code set to the first empty code child, and filling all node codes of the second node code set to the first coding position corresponding to the second empty code child of the empty code child group, and sequentially filling the node codes of the first parent code child except the second node code set to the second empty code child, to generate a child code child group corresponding to the corresponding parent code child group.

[0041] Step 334, repeating the generation of the child code child group corresponding to the corresponding parent code child group to obtain two-point crossover and the child replenishment code body corresponding to the corresponding two replenishment code bodies, and taking all corresponding child replenishment code bodies as the replenishment code bodies completing the current crossover operation.

[0042] Reference Figures 7 to 9 In this embodiment, the tournament selection method is used to select the parent code body, and then the two access codes at the corresponding positions of the access child code bodies at the same position in the parent code body (for example, access code A and access code B in Figure 7 The crossover operation is performed on the two access codes, specifically, a crossover range is randomly generated for the access code A and the access code B, as shown by the dashed line in Figure 7 Then, a parent code child group corresponding empty code child group is generated, including a first empty code child A' and a second empty code child B', and then the node codes (2, 5 / 5, 4) in the crossover range of the access code A and the access code B are copied to the corresponding positions of the first empty code child A' and the second empty code child B', and the access code A is taken as an example, as shown in Figure 8As shown, again, the vacancy position of the first empty encoding sub-A' is filled, the same node encoding in the access encoding sub-B is removed, and then the remaining node encoding of the access encoding sub-B is sequentially filled into the first empty encoding sub-A', thereby forming a complete child access encoding sub-A', as shown in Figure 9 As shown, finally, the child access encoding sub-B' is generated using the same method.

[0043] In this embodiment, exchange mutation is used, and two node encodings in the access encoding sub corresponding to the child transportation path are randomly selected to exchange the access order of the corresponding target supply object, thereby realizing the mutation operation of the inner replenishment encoding body, the access sub-encoding body, and the access encoding sub.

[0044] Step 34, after deciding the candidate replenishment encoding body with the minimum fitness in the at least one candidate replenishment encoding body as the target replenishment encoding body, performing path segmentation on the node sequence corresponding to each access encoding sub of the target replenishment encoding body, and combining at least one target child transportation path corresponding to the generated target path corresponding to the access encoding sub, wherein the target transportation path corresponding to the replenishment decision encoding sub includes the target path.

[0045] Through the above steps 31 to 34, the genetic operation of the inner genetic algorithm of the double-layer genetic algorithm for replenishment path planning is realized to generate the corresponding replenishment transportation path.

[0046] In some embodiments, the fitness corresponding to each multi-dimensional encoding body is determined according to the sub-fitness corresponding to each replenishment decision sub-encoding body, by the following steps: Step 41, obtaining all target transportation paths corresponding to each replenishment decision sub-encoding body and the first replenishment quantity corresponding to each replenishment decision encoding sub, and determining all second replenishment quantities corresponding to each replenishment decision sub-encoding body, wherein the first replenishment quantity is used to represent the total replenishment quantity corresponding to all replenishment quantity encodings corresponding to each replenishment decision encoding sub, and the second replenishment quantity is used to represent the total replenishment quantity corresponding to all replenishment quantity encodings in multiple execution cycles corresponding to each product encoding corresponding to the corresponding replenishment decision sub-encoding body.

[0047] Step 42, after determining the first revenue parameter corresponding to each target product according to the second take-out quantity corresponding to each target product, determining the total sub-revenue parameter corresponding to the replenishment decision sub-encoding body based on the first revenue parameter, wherein the first revenue parameter is determined according to the corresponding sales revenue parameter, the procurement cost parameter, the inventory holding cost parameter, and the out-of-stock cost parameter under the second take-out quantity.

[0048] Step 43, determining a corresponding pickup transportation cost parameter according to the path node corresponding to the target transportation path, and obtaining a transportation cost parameter corresponding to the target transportation path according to the preset departure cost parameter and the pickup transportation cost parameter, and determining a total sub-transportation cost parameter corresponding to the replenishment decision sub-encoding body according to the transportation cost parameter.

[0049] Step 44, performing data processing on the total sub-revenue parameter and the total sub-transportation cost parameter, and based on the data-processed total sub-revenue parameter and the total sub-transportation cost parameter, using Einstein summation convention to solve the corresponding cost parameter to obtain the corresponding sub-fitness, and based on the sub-fitness of all replenishment decision sub-encoding bodies, determining the fitness corresponding to the corresponding multi-dimensional encoding body, wherein the corresponding fitness is used to represent the difference between the total revenue parameter and the total transportation cost parameter, and the data processing includes ndarray array conversion.

[0050] Through the above steps 41 to 44, the fitness of the multi-dimensional encoding body corresponding to the outer replenishment decision is determined.

[0051] In some embodiments, the outer genetic algorithm of the double-layer genetic algorithm performs at least one set of outer genetic evolution operations on the plurality of multi-dimensional encoding bodies participating in the current iteration, including the following steps: Step 51, selecting a preset number of multi-dimensional encoding bodies from the plurality of multi-dimensional encoding bodies participating in the current iteration based on the fitness and a preset selection strategy to obtain candidate multi-dimensional encoding bodies, wherein the selection strategy includes roulette selection and elite reservation strategy.

[0052] In this embodiment, the selection operation combining roulette selection and elite reservation strategy directly reserves the top 2% multi-dimensional encoding bodies with the highest fitness in each generation to the next generation, and the remaining multi-dimensional encoding bodies are selected by roulette selection.

[0053] Step 52, determining replenishment decision sub-encoding bodies associated with the same central warehouse among all replenishment decision sub-encoding bodies corresponding to the candidate multi-dimensional encoding bodies, and performing crossover on the replenishment decision sub-encoding bodies associated with the same central warehouse according to a preset outer crossover operation to generate candidate multi-dimensional encoding bodies that complete the current crossover operation to obtain candidate multi-dimensional encoding bodies.

[0054] In this embodiment, for the three-dimensional structure of the candidate multi-dimensional encoding bodies, product-level crossover operation and execution cycle-level crossover operation are adopted to obtain candidate multi-dimensional encoding bodies, and the selection of each crossover operation is selected with a certain probability.

[0055] To implement the product-level crossover operation, in some optional embodiments, the step 52 of performing crossover on the replenishment decision sub-encoding bodies associated with the same central warehouse according to the preset outer crossover operation includes the following steps: Step 521-1, selecting two replenishment decision sub-codes from all replenishment decision sub-codes associated with the same central warehouse, to obtain a first target sub-code.

[0056] Step 521-2, obtaining a first replenishment quantity code set associated with a target product corresponding to each product code in the first target sub-code, wherein the first replenishment quantity code set includes all replenishment quantity codes of the target product in multiple execution cycles.

[0057] Step 521-3, randomly selecting two first replenishment quantity code sets corresponding to the same product code of the two first target sub-codes, and exchanging all replenishment quantity codes of the two selected first replenishment quantity code sets, to generate a first replenishment decision sub-code corresponding to the replenishment decision sub-codes participating in the current outer cross operation, wherein the candidate multi-dimensional code includes the corresponding first replenishment decision sub-code.

[0058] In this embodiment, for the two first target sub-codes (referring to parent 1 and parent 2 in Figure 10 , the operation exchanges the allocation values of the target products in the same central warehouse and all execution cycles, that is, exchanges the corresponding replenishment quantity codes, that is, exchanges the corresponding first replenishment quantity code sets (referring to a row of replenishment quantity codes in the frame in Figure 10 ), to generate two corresponding first replenishment decision sub-codes (referring to child 1 and child 2 in Figure 11 ).

[0059] To implement the execution cycle level cross operation, referring to Figure 12 and Figure 13 , in some optional embodiments, the step 52 of crossing the replenishment decision sub-codes associated with the same central warehouse according to the preset outer cross operation further includes the following steps: Step 522-1, selecting two replenishment decision sub-codes from all replenishment decision sub-codes associated with the same central warehouse, to obtain a second target sub-code.

[0060] Step 522-2, obtaining a replenishment decision code corresponding to each execution cycle in the second target sub-code.

[0061] Step 522-3, randomly selecting two replenishment decision codes corresponding to the same execution cycle of the two second target sub-codes, and exchanging the two selected replenishment decision codes, to generate a second replenishment decision sub-code corresponding to the replenishment decision sub-codes participating in the current outer cross operation, wherein the candidate multi-dimensional code includes the corresponding second replenishment decision sub-code.

[0062] In this embodiment, for the two second target sub-codes (referring toFigure 12 the same center warehouse and the same target product, exchange the assignment value in the same execution cycle, that is, exchange all the replenishment quantity encodings corresponding to the execution cycle, that is, exchange all the replenishment quantity encodings of the corresponding replenishment decision encoding (see Figure 12 the corresponding column of replenishment quantity encodings in the frame line), to generate two corresponding second replenishment decision encoding bodies (see Figure 11 child 1 and child 2 in the frame line in FIG. 6).

[0063] Step 53, randomly mutate the replenishment quantity encodings of the candidate multi-dimensional encoding body based on the vehicle capacity constraint to generate the multi-dimensional encoding body after the outer genetic evolution operation.

[0064] In the preferred embodiment of the present application, the replenishment decision encoding of the candidate multi-dimensional encoding body is also subjected to a random mutation operation, and a gene bit (corresponding to a replenishment quantity encoding in the replenishment decision encoding) is mutated. Under the premise that the total replenishment quantity of all sub-transportation paths corresponding to at least one replenishment decision encoding satisfies the vehicle capacity constraint of the corresponding replenishment vehicle, the replenishment quantity encoding of the candidate multi-dimensional encoding body is randomly changed with a certain mutation probability to change the gene value of the gene bit (change the encoding value of the replenishment quantity encoding, that is, change the corresponding replenishment quantity), improve the local search ability of the genetic algorithm, maintain the diversity of the population, and prevent premature phenomena.

[0065] In some embodiments, the predicted sales of the goods include the following steps: Step 61, after executing the replenishment and transportation decisions in the first execution cycle of the previous decision phase, obtaining the current inventory quantity corresponding to each center warehouse, and obtaining the preset characteristic parameters corresponding to each center warehouse, wherein the characteristic parameters at least include one of the following: commodity seasonal attribute parameters, commodity transaction attribute parameters.

[0066] In this embodiment, the experimental data is derived from the regional historical data of 6 kinds of goods of a self-operated e-commerce company, each data set is divided into a training set (80%) and a test set (20%), and 7 selected features include "seasonal attribute", "number of views", "number of added shopping carts", "number of collections", "original price", "discount", "advertising intensity", and 1 column of sales data.

[0067] Step 62, determining the out-of-warehouse quantity corresponding to each center warehouse according to the current inventory quantity and the historical decision demand quantity, wherein the historical decision demand quantity is used to represent the product demand quantity of the replenishment and transportation decisions corresponding to the center warehouse in the first execution cycle of the previous decision phase.

[0068] Step 63, input the corresponding out-of-warehouse quantity of each central warehouse and the corresponding characteristic parameter into the sales prediction model, and output the corresponding goods sales of each central warehouse.

[0069] The prediction and training of the sales prediction model of the embodiment of the present application are described as follows:

[0070] In the present embodiment, PSO is applied to optimize two key hyperparameters of the LSTM model: the number of hidden layers and the number of neurons.

[0071] In some preferred embodiments, the PSO optimization of the LSTM hyperparameters includes the following steps: Step 1: import the preprocessed data set, and divide the data set into a training set and a test set.

[0072] Step 2: build a variable-structure LSTM model, and use a loop to dynamically add the number of hidden layers.

[0073] Step 3: initialize the PSO optimization model, randomly generate the number of hidden layers and the number of neurons of each particle to initialize the particle position.

[0074] Step 4: pass the position value of each particle to the variable-structure LSTM model, thereby constructing multiple LSTM models with different structures.

[0075] Step 5: train each LSTM model using the training data set, and calculate the loss function value of each model on the validation set.

[0076] Step 6: pass the loss function value to the PSO model, and take it as the fitness function value of each particle.

[0077] Step 7: update pbest and gbest.

[0078] Step 8: update the speed and position of each particle.

[0079] Step 9: check whether the maximum number of PSO iterations is reached, if not, return to step 4; if reached, output the LSTM model with the optimal structure.

[0080] Step 10: input the inventory quantity of the central warehouse as input data into the trained optimal structure LSTM model, and output the goods sales prediction result.

[0081] The corresponding decision result and the mathematical function model constructed based on the Bi-GA multi-central warehouse joint replenishment and transportation decision of the embodiment of the present application are described as follows:

[0082] Firstly, the following conditions are assumed: (1) the cost of procurement and transportation occurs in the product-to-arrival time period; (2) the demand of each central warehouse is determined by sales forecast, thus is regarded as known in the model; (3) each target supply object provides multiple target products, and each target product is provided by only one target supply object; (4) one execution cycle is 1 day; (5) the replenishment lead time is set to 1 day, i.e. the time for the target product to be transported from the target supply object to the central warehouse is 1 day; (6) the procurement price of the target product adopts a three-interval discount method, i.e. different price discounts are set according to different total procurement quantities of each target product in all central warehouses; (7) the initial inventory quantity of each target product in the central warehouse is set to 0; (8) the central warehouse is allowed to be out of stock, and a shortage penalty cost is set; (9) no horizontal allocation between central warehouses is considered; (10) each target supply object can be accessed by vehicles from only one central warehouse in each execution cycle at most; (11) each transportation vehicle takes a unique designated central warehouse as the starting point, and must return to the original warehouse after performing the pickup task, and the vehicle capacity resources of different warehouses operate independently; (12) heterogeneous vehicle types are used for pickup, and different vehicle types have different start-up costs, unit transportation costs and capacity limits.

[0083] In this embodiment, the target is to maximize the profit of the operation process, i.e. the total sales revenue D minus the procurement cost C1, the transportation cost C2, the inventory holding cost C3 and the shortage cost C4, specifically, the sales revenue D, ; the procurement cost C1, ; the transportation cost C2, ; the inventory holding cost C3, , wherein, ; the shortage cost C4, .

[0084] In this embodiment, the complete mathematical expression of the nonlinear integer programming model constructed for the joint replenishment problem (JRP-T) with flexible replenishment strategy and transportation is as follows: Max Z = D - C1 - C2 - C3 - C4.

[0085] Constraint conditions: 1. The total procurement quantity of each target product in each cycle is constrained by the formula: ; 2. The inventory update formula: ; 3. The total volume of the target product replenished by the wth central warehouse in the tth execution cycle at the target supply object s , ; 4. The flow balance constraint ; ; ; 5. Access frequency limit, ; 6. Vehicle quantity limit ; ; 7. MTZ constraint with capacity constraint , , , , ; 8. Consistency of replenishment and transportation behavior , ; , 9. Inventory capacity constraint .

[0086] The embodiment also provides a Bi-GA-based multi-center warehouse joint replenishment and transportation decision device, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the terms "module", "unit", "sub-unit", etc. can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware is also possible and is contemplated.

[0087] Figure 14 is a structural block diagram of a Bi-GA-based multi-center warehouse joint replenishment and transportation decision device according to the embodiment of the application, as shown in Figure 14 The device comprises a determination module 141, an encoding module 142, and a processing module 143, wherein

[0088] The determination module 141 is configured to determine current demand quantities of a plurality of target products to be replenished in a plurality of execution periods corresponding to a decision stage after the current in a plurality of center warehouses, wherein the current demand quantity is a predicted sales quantity of goods obtained according to a current inventory quantity of the plurality of center warehouses by using a preset sales prediction model, the sales prediction model is a model based on PSO-ALSTM training, and the plurality of execution periods corresponding to the decision stage after the current is generated by rolling a plurality of execution periods corresponding to a last decision stage by one execution period.

[0089] The encoding module 142 is coupled to the determining module 141, and is configured to encode and initialize the central warehouses, the execution cycles, the target products and the current demand quantities according to a preset encoding mode, to generate a plurality of multi-dimensional encoding bodies, wherein the multi-dimensional encoding bodies include a plurality of replenishment decision sub-encoding bodies, one replenishment decision sub-encoding body is associated with one central warehouse, and the replenishment decision sub-encoding body includes a plurality of replenishment decision encoding sub-bodies for representing replenishment decisions of the corresponding central warehouse for the plurality of target products in each execution cycle, and the replenishment decision at least includes a replenishment quantity of the target product.

[0090] The processing module 143 is coupled to the encoding module 142, and is configured to perform inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration based on the replenishment decision sub-encoding bodies corresponding to the plurality of multi-dimensional encoding bodies by using a preset double-layer genetic algorithm, until a target multi-dimensional encoding body is generated, and all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional encoding body are taken as the decision result, wherein the target replenishment path is generated by performing replenishment path planning on the replenishment decision encoding sub-bodies of the replenishment decision sub-encoding body of the target multi-dimensional encoding body by using an inner-layer genetic algorithm of the double-layer genetic algorithm.

[0091] The embodiment also provides a service platform, including a memory and a processor, the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any of the above method embodiments.

[0092] Optionally, the system can 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.

[0093] Optionally, in the embodiment, the processor can be configured to execute the following steps by using the computer program:

[0094] S1, determining current demand quantities of a plurality of target products to be replenished in a plurality of execution cycles corresponding to a decision stage after the current in a plurality of central warehouses, wherein the current demand quantity is a predicted sales quantity of goods obtained according to a current inventory quantity of the plurality of central warehouses by using a preset sales prediction model, the sales prediction model is a model based on PSO-ALSTM, and the plurality of execution cycles corresponding to the decision stage after the current are generated by rolling the plurality of execution cycles corresponding to the last decision stage by one execution cycle.

[0095] S2, encode and initialize the central warehouse, execution cycle, target product and current demand according to a preset encoding mode to generate a plurality of multi-dimensional encoding bodies, wherein the multi-dimensional encoding bodies comprise a plurality of replenishment decision sub-encoding bodies, one replenishment decision sub-encoding body is associated with one central warehouse, the replenishment decision sub-encoding body comprises a plurality of replenishment decision encoding sub-bodies for representing replenishment decisions of the corresponding central warehouse for replenishing a plurality of target products in each execution cycle, the replenishment decision at least comprises a replenishment quantity of the target product.

[0096] S3, using a preset double-layer genetic algorithm, based on the replenishment decision sub-encoding bodies corresponding to the plurality of multi-dimensional encoding bodies, performing inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration until a target multi-dimensional encoding body is generated, and taking all replenishment decisions and all target replenishment paths corresponding to the target multi-dimensional encoding body as the decision result, wherein the target replenishment path is generated by using an inner-layer genetic algorithm of the double-layer genetic algorithm to perform replenishment path planning on the replenishment decision encoding sub-bodies of the replenishment decision sub-encoding bodies of the target multi-dimensional encoding body.

[0097] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation manners, and this embodiment will not be described here.

[0098] In addition, in combination with the multi-central-warehouse joint replenishment and transportation decision method based on Bi-GA in the above embodiments, the present embodiment can provide a storage medium for implementation. The storage medium has a computer program stored thereon; the computer program is executed by a processor to implement any one of the multi-central-warehouse joint replenishment and transportation decision methods based on Bi-GA in the above embodiments.

[0099] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any manner, and in order to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present disclosure.

[0100] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A multi-center warehouse joint replenishment and transportation decision-making method based on Bi-GA, characterized in that, include: Determine the current demand for various target products to be replenished in multiple execution cycles corresponding to a decision stage following the current one in multiple central warehouses. The current demand is the sales volume of goods predicted by a preset sales forecasting model based on the current inventory of the multiple central warehouses. The sales forecasting model is a model trained on PSO-ALSTM. The multiple execution cycles corresponding to the decision stage following the current one are generated by rolling over the multiple execution cycles corresponding to the previous decision stage by one execution cycle. The central warehouse, the execution cycle, the target product, and the current demand are encoded and initialized according to a preset encoding method to generate multiple multi-dimensional encoding bodies. Each multi-dimensional encoding body includes multiple replenishment decision sub-encoding bodies. Each replenishment decision sub-encoding body is associated with a central warehouse. Each replenishment decision sub-encoding body includes multiple replenishment decision encoding bodies used to represent the replenishment decisions made by the corresponding central warehouse for various target products in each execution cycle. The replenishment decision includes at least the replenishment quantity of the target product. Using a pre-defined two-layer genetic algorithm, based on the replenishment decision sub-code bodies corresponding to multiple multidimensional code bodies, an inner layer of replenishment path planning decision iteration and an outer layer of replenishment decision iteration are performed until a target multidimensional code body is generated. All replenishment decisions and all target replenishment paths corresponding to the target multidimensional code body are taken as decision results. The target replenishment path is generated by using the inner layer genetic algorithm of the two-layer genetic algorithm to plan replenishment paths for the replenishment decision sub-code bodies of the target multidimensional code body.

2. The method according to claim 1, characterized in that, Using a pre-defined two-layer genetic algorithm, based on the replenishment decision sub-encoders corresponding to multiple multidimensional encoders, the algorithm performs inner-layer replenishment path planning decision iteration and outer-layer replenishment decision iteration, including: Using the inner layer genetic algorithm of the two-layer genetic algorithm, replenishment path planning is performed on all replenishment decision codes corresponding to the multiple multidimensional codes participating in the current replenishment decision iteration, to obtain the target transportation path corresponding to each replenishment decision code. After determining the fitness of each multidimensional code based on the sub-fitness of each replenishment decision sub-code, the outer genetic algorithm of the two-layer genetic algorithm is used to perform at least one set outer genetic evolution operation on the multiple multidimensional codes participating in the current iteration, generating multiple multidimensional codes that have completed the current replenishment decision iteration. The sub-fitness is determined based on the replenishment transportation cost of all target transportation paths corresponding to the replenishment decision sub-code and the revenue parameter corresponding to the replenishment decision sub-code. Based on the multiple multidimensional codes that have completed the current replenishment decision iteration, the steps of planning and generating the corresponding target transportation path using the corresponding inner genetic algorithm, determining the fitness of the corresponding multidimensional code, and performing the set outer genetic evolution operation using the corresponding outer genetic algorithm are repeatedly executed until multiple multidimensional codes with fitness greater than the fitness threshold are generated. Among the multiple multidimensional encoders with fitness greater than the fitness threshold, the multidimensional encoder with the highest fitness is selected to obtain the target multidimensional encoder.

3. The method according to claim 2, characterized in that, Using the inner layer genetic algorithm of the two-layer genetic algorithm, replenishment path planning is performed on all replenishment decision codecs corresponding to the multiple multidimensional codecs participating in the current replenishment decision iteration, including: Obtain the replenishment decision sub-encoder corresponding to each of the multi-dimensional encoding bodies participating in the current replenishment decision iteration, and determine all the replenishment decision encoding sub-encoders of each of the replenishment decision sub-encoders, wherein the replenishment decision encoding sub-encoders include multiple product codes and replenishment quantity codes, the replenishment quantity code is used to characterize the replenishment quantity of a target product in one execution cycle, and one product code is associated with one target supply object; Each replenishment decision code uses the target supply object associated with all product codes as node codes to generate a node sequence. Each node sequence is then used as an access code to generate an access sub-code body corresponding to the replenishment decision code body. A replenishment code body comprising multiple access sub-code bodies is also generated, where each replenishment code body corresponds to one multi-dimensional code body. Based on vehicle capacity constraints, the node sequence corresponding to the access code body is path-segmented to generate at least one sub-transport path. Each candidate transport path corresponding to the access code body includes at least one sub-transport path. The replenishment quantity for node replenishment at each node code is determined based on the replenishment quantity codes corresponding to all product codes associated with a target supply object. After determining the replenishment cost parameters corresponding to the candidate transportation path based on the sub-replenishment cost parameters corresponding to the sub-transportation path, the decision fitness corresponding to the replenishment code body is determined based on the replenishment cost parameters of the access sub-code body. Based on the decision fitness, an inner-layer genetic algorithm is used to iterate the access sub-code bodies of multiple replenishment codes with a set inner-layer genetic evolution operation until at least one candidate replenishment code body with a decision fitness less than a set fitness value is generated. The inner-layer genetic evolution operation includes the following operations: tournament selection operation, two-point crossover operation, and exchange mutation operation. After selecting the candidate replenishment code with the lowest decision fitness among at least one candidate replenishment code as the target replenishment code, the node sequence corresponding to each access code of the target replenishment code is path segmented, and at least one target sub-transport path generated accordingly is combined into the target path corresponding to the access code, wherein the target transport path corresponding to the replenishment decision code includes the target path.

4. The method according to claim 3, characterized in that, Performing a defined inner-layer genetic evolution operation on the access sub-codemas of the plurality of replenishment codemas, including: After selecting two replenishment codes from a plurality of replenishment codes as parent codes, two access sub-codes located at the same sorting position are selected from all access sub-codes of the two parent codes, and at least one access code with the same sorting position is randomly selected from all access codes of the two corresponding access sub-codes, to obtain at least one set of parent code sub-groups, and to generate an empty code sub-group corresponding to the corresponding parent code sub-group; From all the node codes corresponding to the first parent code and the second parent code of the parent code subgroup, a preset number of node codes located at the same first coding position are randomly selected to obtain the first node code set and the second node code set. All the node codes of the first node code set are respectively filled into the first code bit corresponding to the first empty code bit of the empty code subgroup, and the node codes of the second parent code subgroup other than the first node code set are sequentially filled into the first empty code bit. All the node codes of the second node code set are respectively filled into the first code bit corresponding to the second empty code bit of the empty code subgroup, and the node codes of the first parent code subgroup other than the second node code set are sequentially filled into the second empty code bit, so as to generate a child code subgroup corresponding to the corresponding parent code subgroup. Repeatedly generate the child code subgroup corresponding to the corresponding parent code subgroup to obtain the child replenishment code body that completes the two-point cross and corresponds to the two replenishment code bodies, and use all the corresponding child replenishment code bodies as the replenishment code body that completes the current cross operation.

5. The method according to claim 3, characterized in that, Based on the sub-fitness corresponding to each replenishment decision sub-code body, determine the fitness corresponding to each multidimensional code body, including: Obtain all target transportation routes corresponding to each replenishment decision sub-code body and the first replenishment quantity corresponding to each replenishment decision sub-code body, and determine all second replenishment quantities corresponding to each replenishment decision sub-code body. The first replenishment quantity is used to represent the total replenishment quantity corresponding to all replenishment quantity codes corresponding to each replenishment decision sub-code body, and the second replenishment quantity is used to represent the total replenishment quantity corresponding to all replenishment quantity codes of each product code corresponding to the corresponding replenishment decision sub-code body in multiple execution cycles. After determining the first revenue parameter corresponding to each target product based on the second order quantity corresponding to each target product, the total sub-revenue parameter corresponding to the replenishment decision sub-code body is determined based on the first revenue parameter, wherein the first revenue parameter is determined based on the sales revenue parameter, procurement cost parameter, inventory holding cost parameter and stockout cost parameter corresponding to the second order quantity; Based on the path node corresponding to the target transportation route, determine the corresponding pickup transportation cost parameter, and based on the preset departure cost parameter and the pickup transportation cost parameter, obtain the transportation cost parameter corresponding to the target transportation route, and based on the transportation cost parameter, determine the total sub-transportation cost parameter corresponding to the replenishment decision sub-code body; The total sub-revenue parameter and the total sub-transportation cost parameter are processed, and based on the processed total sub-revenue parameter and the total sub-transportation cost parameter, the corresponding cost parameter is solved using the Einstein summation convention method to obtain the corresponding sub-fitness. Based on the sub-fitness of all the replenishment decision sub-encoders, the fitness of the corresponding multidimensional code is determined. The corresponding fitness is used to characterize the difference between the total revenue parameter and the total transportation cost parameter. The data processing includes ndarray array conversion.

6. The method according to claim 2, characterized in that, Using the outer layer genetic algorithm of the two-layer genetic algorithm, at least one predetermined outer layer genetic evolution operation is performed on the multiple multidimensional codes participating in the current iteration, including: From the multiple multidimensional encoders participating in the current iteration, a preset number of multidimensional encoders are selected based on the fitness and a preset selection strategy to obtain candidate multidimensional encoders, wherein the selection strategy includes roulette wheel selection and elite retention strategy. Among all the replenishment decision sub-code bodies corresponding to the candidate multidimensional code bodies, the replenishment decision sub-code bodies associated with the same central warehouse are determined, and the replenishment decision sub-code bodies associated with the same central warehouse are cross-crossed according to the preset outer cross-cross operation to generate the candidate multidimensional code bodies that have completed the current cross-cross operation, thus obtaining the candidate multidimensional code bodies. The replenishment quantity code of the candidate multidimensional code is subjected to random mutation based on vehicle capacity constraints to generate the multidimensional code that completes the current outer layer genetic evolution operation.

7. The method according to claim 6, characterized in that, For the replenishment decision sub-code bodies associated with the same central warehouse, cross-linking is performed according to a preset outer cross-linking operation, including: From all the replenishment decision sub-code bodies associated with the same central warehouse, select two replenishment decision sub-code bodies to obtain the first target sub-code body; In the first target sub-encoding body, a first replenishment quantity encoding set associated with the target product corresponding to each product code is obtained, wherein the first replenishment quantity encoding set includes all the replenishment quantity codes of the target product in multiple execution cycles; Randomly select two first replenishment quantity code sets corresponding to the same product code of the first target sub-code bodies, and exchange all the replenishment quantity codes of the two selected first replenishment quantity code sets to generate a first replenishment decision sub-code body corresponding to the replenishment decision sub-code body participating in the current outer layer cross operation, wherein the candidate multidimensional code body includes the corresponding first replenishment decision sub-code body.

8. The method according to claim 6, characterized in that, For the replenishment decision sub-code bodies associated with the same central warehouse, cross-linking is performed according to a preset outer cross-linking operation, including: From all the replenishment decision sub-code bodies associated with the same central warehouse, select two replenishment decision sub-code bodies to obtain the second target sub-code body; In the second target sub-encoding body, the replenishment decision encoding sub-encoding corresponding to each execution cycle is obtained; Two replenishment decision codes corresponding to the same execution cycle of two second target sub-code bodies are randomly selected, and the two selected replenishment decision codes are exchanged to generate a second replenishment decision code body corresponding to the replenishment decision code body participating in the current outer layer cross operation, wherein the candidate multidimensional code body includes the corresponding second replenishment decision code body.

9. The method according to claim 1, characterized in that, Predicting the sales volume of the goods includes: After executing the replenishment and transportation decisions of the first execution cycle of the previous decision-making stage, the current inventory of multiple central warehouses is obtained, and the preset characteristic parameters corresponding to each central warehouse are obtained, wherein the characteristic parameters include at least one of the following: commodity seasonal attribute parameters and commodity transaction attribute parameters; Based on the current inventory level and historical decision demand, the outbound quantity corresponding to each central warehouse is determined, wherein the historical decision demand is used to characterize the product demand for replenishment goods decided by the central warehouse in the first execution cycle corresponding to the previous decision stage. The outbound quantity and corresponding characteristic parameters of each central warehouse are input into the sales forecast model, and the sales volume of goods corresponding to the current demand of each central warehouse is output.

10. A service platform, comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the steps of the Bi-GA-based multi-center warehouse joint replenishment and transportation decision-making method as described in any one of claims 1 to 9.

Images