CN117709852A

CN117709852A - Warehouse management method, electronic equipment and storage medium

Info

Publication number: CN117709852A
Application number: CN202311607850.3A
Authority: CN
Inventors: 郭瑞
Original assignee: Beijing Kuangshi Robot Technology Co Ltd
Current assignee: Beijing Kuangshi Robot Technology Co Ltd
Priority date: 2023-11-28
Filing date: 2023-11-28
Publication date: 2024-03-15

Abstract

The embodiment of the application provides a warehouse management method, electronic equipment and storage medium, wherein the method comprises the following steps: determining a first coding vector of a target SKU to be put in storage; determining a first SKU affine value between the target SKU and each stored SKU according to the first encoding vector and a second encoding vector of each stored SKU in the warehouse, wherein the first SKU affine reflects the probability of the target SKU and the stored SKU in the same order; determining a target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU; obtaining estimated library information of a target SKU in a future target time period; wherein, the estimated ex-warehouse information is determined based on the historical order corresponding to the target SKU; and determining the target storage position of the target SKU in the target storage area according to the estimated library outlet information of the target SKU. The embodiment of the application can improve the warehouse-in efficiency and the whole warehouse-out efficiency of the warehouse.

Description

Warehouse management method, electronic equipment and storage medium

Technical Field

The embodiment of the application relates to the technical field of logistics, in particular to a warehouse management method, electronic equipment and a storage medium.

Background

With the development of e-commerce technology, the scale of a warehouse is continuously enlarged, the types of articles stored in the warehouse are more and more, and the operation efficiency of warehouse entry and warehouse exit is affected by how the articles of the types are subjected to warehouse entry and warehouse exit.

SKU (Stock Keeping Unit, minimum stock keeping unit) is an important dimension in the prior art, whether in-warehouse or out-warehouse. Typically, the warehouse entry and exit operations are guided based on the metrics represented by the individual SKUs. However, the individual SKUs are used for guiding the warehouse-in and warehouse-out operation, and the relationship among different SKUs is not considered, so that the warehouse-in and warehouse-out operation efficiency is affected to a certain extent.

Disclosure of Invention

In view of the foregoing, embodiments of the present application are presented to provide a warehouse management method, an electronic device, and a storage medium that overcome or at least partially solve the foregoing problems.

According to a first aspect of embodiments of the present application, there is provided a warehouse management method, including:

determining a first coding vector of a target SKU to be put in storage;

determining a first SKU affine value between a target SKU and each stored SKU according to the first encoding vector and a second encoding vector of each stored SKU in a warehouse, wherein the first SKU affine value reflects the probability of the target SKU and the stored SKU in the same order;

Determining a target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU;

obtaining estimated library information of the target SKU in a future target time period; wherein the estimated ex-warehouse information is determined based on the historical order corresponding to the target SKU;

and determining a target storage position of the target SKU in the target storage area according to the estimated library outlet information of the target SKU.

According to a second aspect of embodiments of the present application, there is provided an electronic device, including: a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor implements the warehouse management method as described in the first aspect.

According to a third aspect of embodiments of the present application, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the warehouse management method according to the first aspect.

According to a fourth aspect of embodiments of the present application, there is provided a computer program product comprising a computer program or computer instructions which, when executed by a processor, implement the warehouse management method of the first aspect.

According to the warehouse management method, the electronic device and the storage medium, through determining the first coding vector of the target SKU to be warehoused, according to the first coding vector and the second coding vector of each stored SKU in the warehouse, determining the first SKU affine value between the target SKU and each stored SKU, according to the first SKU affine value and the storage area of each stored SKU, determining the target storage area of the target SKU, obtaining the estimated ex-warehouse information of the target SKU in a future target time period, according to the estimated ex-warehouse information, determining the target storage position of the target SKU in the target storage area, fully considering the relation between different SKUs through the first SKU affine value when determining the target storage area of the target SKU, further guiding the warehouse-in of the target SKU to be warehoused based on the first SKU affine value, establishing the relation between the target SKU and the stored SKU, and improving the warehouse-in efficiency and improving the whole ex-warehouse efficiency.

The foregoing description is only an overview of the technical solutions of the present application, and may be implemented according to the content of the specification in order to make the technical means of the present application more clearly understood, and in order to make the above-mentioned and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.

Drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application.

FIG. 1 is a flow chart of steps of a warehouse management method provided in an embodiment of the present application;

FIG. 2 is a flow chart of a training process of a word vector model in an embodiment of the present application;

FIG. 3 is a schematic diagram of the structure of a Skip-gram in an embodiment of the present application;

FIG. 4 is a schematic illustration of a self-climbing robotic warehouse in an embodiment of the present application;

FIG. 5 is a flowchart of steps of a warehouse management method provided in an embodiment of the present application;

fig. 6 is a block diagram of a warehouse management device according to an embodiment of the present application;

fig. 7 is a block diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

Along with the development of intelligent technologies such as the Internet of things, artificial intelligence and big data, the demands of transformation and upgrading of traditional logistics industry by utilizing the intelligent technologies are stronger, and intelligent logistics (Intelligent Logistics System) becomes a research hotspot in the logistics field. The intelligent logistics utilizes the Internet of things devices and technologies such as artificial intelligence, big data, various information sensors, radio frequency identification technology, global Positioning System (GPS) and the like, is widely applied to basic movable links such as transportation, storage, distribution, packaging, loading and unloading of materials, information service and the like, and realizes intelligent analysis decision, automatic operation and high-efficiency optimization management of the material management process. The internet of things technology comprises sensing equipment, RFID technology, laser infrared scanning, infrared sensing identification and the like, and can effectively connect materials in logistics with a network, monitor the materials in real time, sense environmental data such as humidity and temperature of a warehouse and guarantee the storage environment of the materials. All data in the logistics can be perceived and collected through a big data technology, the data are uploaded to an information platform data layer, operations such as filtering, excavating, analyzing and the like are carried out on the data, and finally accurate data support is provided for business processes (such as links of transportation, warehousing, access, picking, packaging, sorting, warehouse-out, inventory, distribution and the like). The application direction of artificial intelligence in logistics can be broadly divided into two types: 1) The intelligent equipment such as an unmanned truck, an AGV, an AMR, a forklift, a shuttle, a stacker, an unmanned delivery vehicle, an unmanned plane, a service robot, a mechanical arm, an intelligent terminal and the like which are energized by the AI technology is used for replacing part of manpower; 2) The manual efficiency is improved through a software system driven by technologies or algorithms such as computer vision, machine learning, operation optimization and the like, such as a transportation equipment management system, warehouse management, equipment scheduling system, order distribution system and the like. With the research and advancement of smart logistics, the technology has expanded applications in numerous fields, such as retail and electronics, tobacco, medicine, industrial manufacturing, footwear, textiles, food, etc.

Fig. 1 is a flowchart of steps of a warehouse management method according to an embodiment of the present application, as shown in fig. 1, the method may include:

step 101, determining a first encoding vector of a target SKU to be put in storage.

And encoding the target SKU to be put into storage to obtain a first encoding vector of the target SKU. The target SKU to be put in storage can be encoded through a word vector model, and the target SKU to be put in storage can be encoded through other neural network models.

In one embodiment of the present application, the determining the first encoding vector of the target SKU to be binned includes:

when a single-hot representation vector of a target SKU is used as an input of a word vector model, a hidden layer output vector of the word vector model is used as a first coding vector of the target SKU; or,

and when the single-hot representation vector is used as the output of the word vector model, using the hidden layer input vector of the word vector model as the first coding vector of the target SKU.

The word vector model may include CBOW (continuous bag of words, continuous word bag model) or Skip-gram, among others. The basic idea of CBOW is to predict a word itself given the context of that word (i.e., other words within the window). CBOW is a method of training a neural network using word embedding, where a context is represented by multiple words of a given target word. Skip-gram is the context word that is predicted given the target word.

The word vector model may be trained based on individual SKUs in the historical orders. Based on different forms of the word vector model, the target SKU may be input or output, such that a first encoding vector for the target SKU may be determined based on the network weights of the trained word vector model and the target SKU. For example, when the word vector model is CBOW, the target SKU is taken as the output of the word vector model; when the word vector model is Skip-gram, the target SKU is taken as the input of the word vector model.

The sequential positions of all SKUs (SKUs of all articles to be stored in the warehouse) are predetermined, and then, based on the positions of the target SKU in all SKUs, a one-hot (one-hot) representation vector of the target SKU can be obtained, based on the one-hot representation vector of the target SKU and a word vector model, a hidden layer output vector or a hidden layer input vector in the word vector model can be determined, and the hidden layer output vector or the hidden layer input vector is determined as a first encoding vector of the target SKU, that is, the product of a weight matrix between the target SKU and the hidden layer and the one-hot representation vector of the target SKU is determined as the first encoding vector of the target SKU. For example, when the word vector model is CBOW, the one-hot representation vector of the target SKU is taken as output, a product of the weight matrix between the output layer and the hidden layer and the one-hot representation vector of the target SKU is determined, that is, the hidden layer input vector of the word vector model, and the hidden layer input vector is determined as the first encoding vector of the target SKU; when the word vector model is Skip-gram, the one-hot representation vector of the target SKU is used as input, the product of the weight matrix between the input layer and the hidden layer and the one-hot representation vector of the target SKU is determined, namely the hidden layer output vector of the word vector model, and the hidden layer output vector is determined to be the first coding vector of the target SKU. By taking the hidden layer output vector or the hidden layer input vector in the word vector model determined based on the one-hot representation vector of the target SKU as the first encoding vector of the target SKU, the target SKU can be uniquely represented by the first encoding vector.

Step 102, determining a first SKU affine value between the target SKU and each stored SKU according to the first encoding vector and a second encoding vector of each stored SKU in the warehouse, wherein the first SKU affine value reflects the probability that the target SKU and the stored SKU occur in the same order.

Where the stored SKU is the SKU corresponding to the stored container in the warehouse. The storage area may be a vertical roadway in a self-climbing robot warehouse, or may be two side shelves corresponding to a aisle in a multi-container simultaneous-fetching warehouse (e.g., a multi-tier robot warehouse).

And encoding the stored SKUs to obtain second encoding vectors of the stored SKUs. Illustratively, the second encoded vector of a stored SKU may be determined by a word vector model, i.e., in determining the second encoded vector of a stored SKU, multiplying a weight matrix between a hidden layer in the word vector model and the stored SKU by a one-hot representation vector of the SKU to obtain the second encoded vector of the stored SKU. Affine relationship between two SKUs, i.e., SKU affine value between two SKUs, can be described by a spatial distance, which can be expressed as Affinity (i, j) =dist (V (SKU) ⁱ )*V(SKU ^j ) Dist represents distance, V (SKU) ⁱ ) The code vector representing the ith (the position of this SKU among all SKUs), V (SKU) ^j ) The first SKU affine value between the target SKU and each stored SKU can be obtained by substituting the first encoding vector of the target SKU and the second encoding vector of each stored SKU into the above formula, respectively, representing the encoding vector of the jth SKU (the position of the SKU among all SKUs). The smaller the affine value of the first SKU between the target SKU and the stored SKU, the greater the probability that the target SKU and the stored SKU appear in the same order; the greater the affine value of the first SKU between the target SKU and the stored SKU, the less likely the target SKU will appear in the same order as the stored SKU. The first SKU affine value is an advanced method for evaluating the relationship between the SKUs, can represent the intimate relationship between the target SKU and the stored SKU, can obtain a more easily-expressed mode to express the hidden relationship of the two SKUs under a certain condition, and can further formulate a warehousing strategy according to the first SKU reflection value to guide the warehousing business and improve the warehousing efficiency.

Step 103, determining a target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU.

A target holding area for the target SKU is determined based on the first SKU affine value, the storage area of each stored SKU in the warehouse and the storage characteristics of the warehouse. For example, when the storage characteristics of the warehouse are that the same or similar SKUs are stored adjacently (the warehouse with multiple containers can be taken at one time), determining a storage area of the stored SKUs with smaller affine values of the first SKU, and determining a storage area which is closer to the storage area as a target storage area of the target SKU; when the storage characteristics of the warehouse are the same or similar SKUs in scattered storage, a storage area of the stored SKUs with smaller affine values of the first SKU can be determined, and a storage area far from the storage area is determined as a target storage area of the target SKU.

104, obtaining estimated ex-warehouse information of the target SKU in a future target time period; the estimated ex-warehouse information is determined based on the historical order corresponding to the target SKU.

The estimated ex-warehouse information may be an estimated ex-warehouse amount or an estimated flow rate.

Determining the estimated ex-warehouse quantity of the target SKU in a future target time period based on the historical order comprising the target SKU, and taking the estimated ex-warehouse quantity as estimated ex-warehouse information; or after determining the estimated delivery amount of the target SKU, determining the estimated transfer rate of the target SKU based on the estimated delivery amount and the stock amount of the target SKU, and determining the estimated transfer rate as estimated delivery information.

And 105, determining a target storage position of the target SKU in the target storage area according to the estimated ex-warehouse information.

The higher the estimated ex-warehouse information is, the more easily ex-warehouse storage in the target area, for example, storage close to the picking site is determined as target storage, the lower the estimated ex-warehouse information is, and storage far away from the picking site in the target area is determined as target storage. The corresponding storage bit range can be preset according to the interval range of the estimated delivery information, further, the estimated delivery information is estimated based on the estimated delivery information of the target SKU, the target storage bit in the target area can be selected in the storage bit range corresponding to the estimated delivery information, and the container to be stored can be stored in a warehouse and stored in the target storage bit.

According to the warehouse management method, through determining the first coding vector of the target SKU to be warehoused, according to the first coding vector and the second coding vector of each stored SKU in the warehouse, determining the first SKU affine value between the target SKU and each stored SKU, according to the first SKU affine value and the storage area of each stored SKU, determining the target storage area of the target SKU, obtaining the estimated ex-warehouse information of the target SKU in a future target time period, according to the estimated ex-warehouse information, determining the target storage position of the target SKU in the target storage area, fully considering the relation between different SKUs through the first SKU affine value when determining the target storage area of the target SKU, further guiding the warehouse-in of the target SKU to be warehoused based on the first SKU affine value, establishing the relation between the target SKU and the stored container, and improving the warehouse-in efficiency and improving the whole ex-warehouse efficiency.

FIG. 2 is a flowchart of a training process of a word vector model according to an embodiment of the present application, and as shown in FIG. 2, the training process of the word vector model includes:

step 201, determining a single thermal representation vector corresponding to each order line in each sample order according to the SKU and the SKU number corresponding to each order line in each sample order.

All SKUs were separately single hot (one-hot) coded. All SKUs are numbered one by one, some SKU based on the total number of SKUs (e.g., 10000 total) ^k The corresponding one-hot representation vector may be represented as (0, …,1, … 0), where the vector index bit of 1 is k.

The sample orders may be some orders selected from the historical orders. Each order line in the sample order includes a SKU and the number of SKUs corresponding to the SKU, and the single heat expression vector corresponding to each order line in the sample order may be represented by the SKU corresponding to the order line, where the number of SKUs in all SKUs is denoted by the number of SKUs, and the other number of SKUs is denoted by 0. The ith sample Order of all sample orders may be expressed as Order ⁱ The specimen order includes several ordersRow { (SKU) ^k ,qty ^i,k ) … }, then SKU in the sample order ^k The independent heat expression vector corresponding to the order line is (0, …, qty) ^i,k … 0), qty of which ⁱ ^,k Order for sample Order ⁱ SKU of (b) ^k SKU number of (a).

Step 202, determining a context representation vector corresponding to each order line in each sample order according to the single heat representation vector corresponding to each order line in each sample order.

The context representation vector of one order line in a sample order is the independent heat representation vector (0, …, qty) corresponding to each order line in the sample order except the current order line ^i,k’ … 0), if there are n order lines in an order, the context representation vector for a certain order line may have n-1.

In step 203, the single thermal representation vector and the context representation vector corresponding to each order line in the same sample order are determined as the input-output pair.

And the single-hot representation vector and the context representation vector corresponding to each order row in the same sample order are used as input-output pairs of the training word vector model. For example, when the word vector model is CBOW, the output in one input-output pair is a single hot representation vector corresponding to one order line in the sample order, and the input in the input-output pair is a context representation vector of the order line; when the word vector model is Skip-gram, the input in one input-output pair is a single-hot representation vector corresponding to one order line in the sample order, and the output in the input-output pair is a context representation vector of the order line.

And 204, training the initial word vector model according to the input-output pair to obtain a trained word vector model.

Based on the determined input-output pairs, training an initial word vector model, and adjusting network weights of the word vector model through back propagation. Taking a word vector model as Skip-gram as an example, as shown in fig. 3, the single-hot representation vector of the order line is taken as input, vectors of N hidden units can be obtained at a hidden layer (i.e. a mapping layer), the output vector of the output layer corresponds to the context representation vector of the order line, i.e. the predicted context SKU and SKU number thereof, and the training process uses the back propagation training network weight. N hidden units form a coded vector of the SKU, and the coded vector of the SKU can be obtained by multiplying the weight matrix near the central SKU side by the one-hot representation vector of the SKU.

The word vector model is trained based on the input-output pairs formed by the single-hot expression vector and the context expression vector of each order row in each sample order, so that the word vector model can distinguish the expression forms of different SKUs, and the coded vectors of the different SKUs can be expressed more accurately.

In one embodiment of the present application, the determining the target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU includes: clustering all SKUs according to the first SKU affine value and a second SKU affine value between every two SKUs in the stored SKUs, and determining a target group corresponding to the target SKU; if the storage area corresponding to the target group has empty storage bits, determining the storage area corresponding to the target group as the target storage area; and if the target group is not allocated with a storage area or no empty storage bit exists in the storage area corresponding to the target group, determining the target storage area according to the estimated library outlet information of the target SKU.

In ex-warehouse, if there is a dependency on a pick item (SKU), such as a container in a dense warehouse may be a hindrance to another container, or a person in a manual warehouse needs to pick multiple containers in succession, then the relationship between different SKUs needs to be considered in the warehouse. In a warehouse that is a multi-container simultaneous-fetching (multi-container-at-a-time) warehouse (e.g., a multi-layer bin robotic warehouse or a manual warehouse, etc.), it is desirable to be able to store SKUs in close proximity to each other, so that multiple containers can be fetched at a time and taken out together.

Taking a multi-layer bin robot warehouse as an example, the multi-layer bin robot can take a plurality of bins (containers) to a picking station at a time, and in the cost of a delivery task, the time cost of the robot to the bin taking position and the time cost of the robot to the picking station at last are included. In addition to the distance from the last pick box location to the picking station, minimizing pick box time and minimizing wait time are also important considerations. Since in such warehouse layouts, often aisles can only be passed by one robot at the same time, it is known from the routing strategies commonly used by robots that the box taking time is shortened if the box taking positions are distributed as much as possible on fewer (and close) aisles. In addition, if one robot works in as few aisles as possible in one task, the waiting probability is also reduced, thereby reducing the waiting time.

When the warehouse is a multi-container concurrent warehouse, clustering all SKUs based on a first SKU affine value and a second SKU affine value between every two SKUs in the stored SKUs, and classifying the SKUs with smaller SKU affine values (including the first SKU affine value and the second SKU affine value) into a group; alternatively, all SKUs are clustered in advance based on a second SKU affine value (at this time, the second SKU affine value includes the first SKU affine value) between every two SKUs among all SKUs (all SKUs that need to be stored or already stored), and SKUs with smaller second SKU affine values are classified into one group. And inquiring the group to which the target SKU belongs from all groups obtained by clustering based on the target SKU to obtain the target group to which the target SKU belongs. If the storage area is allocated to the target group and the empty storage bit exists in the storage area corresponding to the target group, determining the storage area corresponding to the target group as the target storage area of the target SKU; if the target group does not allocate a storage area or no empty storage exists in the storage area corresponding to the target group, when the estimated delivery information (the estimated delivery amount or the estimated delivery rate in the future target time period) of the target SKU is large, the storage area closer to the picking station may be determined as the target storage area, and when the estimated delivery information of the target SKU is small, the storage area farther from the picking station may be determined as the target storage area.

The method comprises the steps of clustering each SKU based on the first SKU affine value and the second SKU affine value between every two SKUs in the stored SKUs to obtain a plurality of groups, and determining the target storage area of the target SKU based on the target group to which the target SKU belongs, so that the target storage area is positioned near the storage area of the target group, and therefore simultaneous delivery of each SKU in the target group is facilitated, and delivery efficiency can be improved.

In another embodiment of the present application, the determining the target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU includes: if the target SKU is stored in the warehouse and the empty storage bit exists in the storage area corresponding to the target SKU, determining the storage area corresponding to the target SKU as the target storage area; if the target SKU is not stored in the warehouse or no empty storage bit exists in a storage area corresponding to the target SKU, determining the target stored SKU of which the affine value of the first SKU is smaller than an affine threshold value, and determining the target storage area according to the storage area of the target stored SKU.

When the warehouse is a multi-container simultaneous taking warehouse, if the target SKU is stored in the warehouse and an empty storage position exists in a storage area corresponding to the target SKU, the storage area corresponding to the target SKU can be determined to be the target storage area, and as an example, the target SKU has a stored container in the warehouse and an empty storage position exists in the storage area corresponding to the stored container, the storage area corresponding to the stored container can be determined to be the target storage area of the container to be stored of the target SKU; if the target SKU is not stored in the warehouse, i.e., the target SKU does not have a stored container in the warehouse, or if the target SKU does not have an empty storage location in a storage area corresponding to the target SKU, i.e., the target SKU does have a stored container in the warehouse but does not have an empty storage location in a storage area corresponding to the stored container, determining a target stored SKU having a first SKU affine value less than an affine threshold from all the stored SKUs, determining a target storage area of the container to be stocked based on the storage area of the target stored SKU, determining the storage area of the stored SKU having the minimum first SKU affine value as the target storage area, or determining the storage area of one of the target stored SKUs as the target storage area.

By determining the storage area corresponding to the target SKU as the target storage area or determining the target storage area based on the storage area of the stored SKUs with the affine values smaller than the affine value threshold, SKUs with smaller affine values of the first SKU can be stored adjacently, so that containers corresponding to the SKUs can be conveniently and simultaneously delivered, and delivery efficiency can be improved.

On the basis of the above technical solution, the determining the target storage area according to the storage area of the target stored SKU includes: if the storage area of the target stored SKU has an empty storage bit, determining the storage area of the target stored SKU with the minimum affine value of the first SKU as the target storage area; if the storage area of the target stored SKU does not have empty storage bits, determining the target storage area according to the estimated ex-warehouse information of the target SKU.

When there is only one target stored SKU, if there is an empty storage bit in the storage area of the target stored KSU, the storage area of the target stored SKU may be determined as the target area, and when there are a plurality of target stored SKUs, the storage area of the stored SKU having the empty storage bit may be determined from the storage areas of the plurality of target stored SKUs, and the storage area of the target stored SKU having the smallest affine value of the first SKU in the storage area of the stored SKU having the empty storage bit may be determined as the target storage area; if none of the storage areas of the target stored SKUs has empty storage bits, the target area may be determined based on the estimated shipment information of the target SKUs, that is, when the estimated shipment information of the target SKUs is large, the storage area closer to the picking station may be determined as the target storage area, and when the estimated shipment information of the target SKUs is small, the storage area farther from the picking station may be determined as the target storage area.

When the storage area of the target stored SKU has the empty storage bit, the storage area of the target stored SKU with the empty storage bit and the minimum affine value of the first SKU is determined as the target storage area, so that the target SKU and the stored SKU can be conveniently and simultaneously delivered, the delivery efficiency can be improved, when the storage area of the target stored SKU does not have the empty storage bit, the target storage area is determined according to the estimated delivery information of the target SKU, and the delivery efficiency of a warehouse can be improved.

In another embodiment of the present application, the determining the target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU includes: determining a distance between each candidate region and each storage region for each candidate region in the warehouse, wherein the candidate regions are storage regions with empty storage bits in the warehouse; determining a third SKU affine value between the target SKU and the stored SKUs in the candidate region according to the first SKU affine value and each distance; the third SKU affine value reflects the probability that the target SKU appears in the same order as the stored SKU in the candidate region; determining a delivery index evaluation value corresponding to the candidate region according to the estimated delivery information of the target SKU, the distance between the candidate region and the picking site and the third SKU affine value; and determining the candidate area with the minimum evaluation value of the ex-warehouse index as the target storage area.

The target storage area of the target SKU may be determined heuristically. Firstly, a storage area with empty storage bits in a warehouse is determined as a candidate area, a warehouse-out index evaluation value is determined according to a heuristic mode for each candidate area, and a candidate area with the minimum warehouse-out index evaluation value is determined as a target storage area. When determining the evaluation value of the ex-warehouse index of a candidate area, determining a third SKU affine value between the target SKU and all stored SKUs in the candidate area according to the following formula:

wherein Affinity (SKU, loc) represents a third SKU affine value between the target SKU and all stored SKUs within the candidate region loc, affinity (SKU, other SKU) represents a first SKU affine value between the target SKU (SKU) and stored SKU (other SKU), loc ^m Representing the storage area corresponding to the stored SKU, dist (loc ) ^m ) Storage area loc representing candidate area loc and stored SKU ^m Distance between them.

The third SKU affine value characterizes affine relationships between the target SKU and storage areas of other stored SKUs. Based on the characteristics of the warehouse, determining the warehouse-out index evaluation value corresponding to the candidate region according to the estimated warehouse-out information of the target SKU, the distance between the candidate region and the picking site and the third SKU affine value. If the warehouse is characterized by taking a plurality of containers at a time (such as a multi-layer bin robot warehouse), a plurality of SKUs with smaller affine values of the first SKU need to be stored adjacently, namely, a storage area with smaller affine values of the third SKU is more likely to be determined as a target storage area; if the warehouse is characterized by decentralized storage (e.g., a self-climbing robotic warehouse), then multiple SKUs with smaller affine values of the first SKU need to be stored in a decentralized manner, i.e., the greater the affine value of the third SKU, the more likely it is to be determined as the target storage area.

By adopting a heuristic method to determine the target area, the relation between the target SKU and the storage area of the stored SKU is fully considered, so that the SKUs in the scattered storage warehouse can not be affected when being delivered, and the SKUs with close relation can be delivered simultaneously in the multi-container simultaneous delivery warehouse, thereby improving the delivery efficiency.

In an alternative embodiment, the warehouse is a multi-container concurrent warehouse;

the determining the evaluation value of the delivery index corresponding to the candidate area according to the estimated delivery information of the target SKU, the distance between the candidate area and the picking site and the affine value of the third SKU comprises:

determining a delivery index evaluation value corresponding to the candidate region according to the estimated delivery information of the target SKU, the distance between the candidate region and the picking site and the third SKU affine value, and the following formula:

y＝a*popular(sku)*distance(loc)+b*affinity(sku,loc)

wherein y represents an evaluation value of a delivery index corresponding to the candidate region, potential (SKU) represents estimated delivery information of the target SKU, distance (loc) represents a distance between the candidate region loc and the picking site, affinity (SKU, loc) represents an affine value of the third SKU, and a and b are adjustment coefficients.

In a multi-container concurrent picking warehouse (such as a multi-layer bin robotic warehouse or a manual warehouse), multiple containers with smaller affine values of a first SKU need to be stored adjacently, while a third SKU affine value is determined based on the first SKU affine value, the sum of the third SKU affine value and a picking score is determined as a picking index evaluation value, the picking score is the product of estimated picking information and the distance between a candidate area and a picking site, and the smaller the picking index evaluation value of one candidate area is, the greater the probability that the candidate area is taken as a target storage area is represented. By determining the ex-warehouse index evaluation value and determining the target storage area based on the ex-warehouse index evaluation value in the mode, SKUs with affinity can be simultaneously ex-warehouse in a multi-container co-warehouse, so that the efficiency in ex-warehouse is improved.

In another alternative embodiment, the warehouse is a self-climbing robotic warehouse;

y＝a*popular(sku)*distance(loc)-b*affinity(sku,loc)

The self-climbing robot warehouse is characterized by scattered storage. Fig. 4 is a schematic diagram of a self-climbing robot warehouse in the embodiment of the application, as shown in fig. 4, the self-climbing robot can only climb up to a high level from one level of the current vertical roadway, so that one level is easy to form a bottleneck, and the robot should work on different vertical roadways as much as possible, that is, the same SKU or similar SKUs are required to be stored in a scattered manner.

In the self-climbing robot warehouse, a plurality of SKUs with smaller first SKU affine values should be avoided from being put into the same storage area (vertical roadway), namely a plurality of containers with smaller first SKU affine values need to be stored in a scattered manner, a third SKU affine value is determined based on the first SKU affine value, a difference between a picking score and the third SKU affine value is determined as a picking index evaluation value, the picking score is a product of estimated picking information and a distance between a candidate area and a picking site, and the smaller the picking index evaluation value of one candidate area is, the greater the probability that the candidate area is taken as a target storage area is represented, so that the same SKU or similar SKU can be stored in a scattered manner. By determining the ex-warehouse index evaluation value and determining the target storage area based on the ex-warehouse index evaluation value in the mode, the self-climbing robot warehouse can ensure that all SKUs can not be affected each other when ex-warehouse, so that the efficiency of ex-warehouse is improved.

Fig. 5 is a flowchart of steps of a warehouse management method according to an embodiment of the present application, as shown in fig. 5, the method may include:

step 601, determining a first encoding vector of a target SKU to be binned.

Step 602, determining a first SKU affine value between the target SKU and each stored SKU in the warehouse according to the first encoding vector and a second encoding vector of each stored SKU, wherein the first SKU affine value reflects the probability that the target SKU and the stored SKU occur in the same order.

Step 603, determining a target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU.

Step 604, obtaining estimated ex-warehouse information of the target SKU in a future target time period; the estimated ex-warehouse information is determined based on the historical order corresponding to the target SKU.

Step 605, determining a target storage position of the target SKU in the target storage area according to the estimated ex-warehouse information.

Step 606, determining an order code vector corresponding to each order to be placed according to the third code vector of each SKU in each order to be placed.

In a retail warehouse, hundreds of thousands of orders may be processed at the same time, the order volume is large, and it is certainly time consuming to do a wave-grouping optimization calculation based on the hundreds of thousands of orders (because the problem is NP-hard). In the embodiment of the application, coarse screening and filtering can be performed through the order affine values among orders, and then wave grouping optimization calculation is performed, so that wave grouping optimization efficiency can be improved.

The average vector of the third encoding vectors of each SKU in an order to be placed may be determined as the order encoding vector of the order to be placed, that is, the order encoding vector of an order to be placed is determined by the following formula:

Vector(Order)＝Sum(Vector(sku ^k )/Num ^sku )

wherein Vector (Order) represents an Order encoding Vector of an Order to be placed, vector (sku) ^k ) Third code vector, num, representing the kth SKU in the order to be placed ^sku Indicating the number of SKUs in the order to be placed.

Step 607, determining an affine value of the order between every two to-be-picked orders according to the order encoding vector of each to-be-picked order, where the affine value of the order reflects the probability of picking the two to-be-picked orders together.

The order affine value is the distance between the order encoding vectors of two orders to be placed. The order affine value between each two to-be-placed orders may be determined according to the following formula:

OrderAffinity(i,j)＝Dist(Vector(order i)*Vector(order j))

wherein OrderA affinity (i, j) represents an order affine value between the i-th to-be-placed order and the j-th to-be-placed order, vector (order i) represents an order encoding vector of the i-th to-be-placed order, vector (order j) represents an order encoding vector of the j-th to-be-placed order, and Dist represents a distance.

And step 608, clustering all the orders to be taken out according to the affine value of the orders to obtain a plurality of order classifications.

Clustering the orders to be placed according to the order affine values, aggregating a plurality of orders to be placed with smaller order affine values into one order category, and classifying all the orders to be placed into a plurality of order categories. For example, after the number of the to-be-clustered or the size of the clusters is selected, all the to-be-ex orders are clustered into a plurality of order classifications by iteratively calculating a clustering center and a clustering range based on affine values of the orders.

Step 609, for each order classification, performing wave grouping processing on each order to be delivered in the classification according to the storage position of SKU in each order to be delivered, so as to obtain at least one wave corresponding to the classification, where the wave is used for delivering and picking.

And respectively carrying out wave-grouping optimization processing on each order category, and when carrying out wave-grouping optimization processing on one order category, carrying out wave-grouping processing on each order to be delivered in the order category according to the storage position of the SKU in each order to be delivered in the order category, and dividing the order category into at least one wave order. The individual wave times obtained by the wave grouping process can be used for jointly carrying out ex-warehouse picking.

When wave combination processing is performed on each order to be delivered in one order category, a heuristic method or a meta-heuristic method can be adopted, the wave number is selected while the SKU is positioned in the order to be delivered (the storage position of the SKU in the order to be delivered is determined), and certain time complexity exists. Heuristic common methods include seed and sampling, wherein seed is understood as continuously traversing the order to be taken out, and adding the order to be taken out to the existing wave according to the order to be taken out (namely seeds) in the wave; saving can be understood as constantly merging the wavelet passes. Meta-heuristic can be understood as continuously exploring new solutions (to-be-taken-out orders) in the neighborhood by operators, and defining the solution with the minimum cost in the exploration process. The steps of order wave combining processing are not limited in the embodiment of the present application, and any mode for implementing order wave combining processing may be used in the embodiment of the present application to perform wave combining processing on each order to be taken out in one order classification.

According to the warehouse management method provided by the embodiment, each order to be checked is clustered into a plurality of order classifications according to the affine value of the order between every two orders to be checked, and then wave combination processing is carried out on each order classification, so that compared with the wave combination processing carried out in the whole range of all orders to be checked, the data volume participating in calculation processing is reduced, the wave combination processing time is shortened, the wave combination efficiency is improved, and the warehouse outlet efficiency of each SKU in all orders to be checked is improved.

Order encoding vectors and affine relationships may also be used in order splitting bins. When the order is divided into bins, the orders can be clustered according to order code vectors, and are converted into balance constraint and affine (affinity) loss, so that a coarse-strength order dividing bin result is formed, and fine adjustment is performed in the coarse-granularity order dividing bin result, so that a final order dividing bin result is obtained.

It should be noted that, for simplicity of description, the method embodiments are shown as a series of acts, but it should be understood by those skilled in the art that the embodiments are not limited by the order of acts described, as some steps may occur in other orders or concurrently in accordance with the embodiments. Further, those skilled in the art will appreciate that the embodiments described in the specification are all preferred embodiments and that the acts referred to are not necessarily required by the embodiments of the present application.

Fig. 6 is a block diagram of a warehouse management device according to an embodiment of the present application, and as shown in fig. 6, the warehouse management device may include:

an encoding module 701, configured to determine a first encoding vector of a target SKU to be put in storage;

a first SKU affine value determination module 702 configured to determine a first SKU affine value between the target SKU and each stored SKU in the warehouse according to the first encoding vector and a second encoding vector of each stored SKU, where the first SKU affine value reflects a probability that the target SKU and the stored SKU occur in the same order;

a target area determining module 703, configured to determine a target storage area of the target SKU according to the affine value of the first SKU and the storage area of each stored SKU;

the estimated ex-warehouse information acquisition module 704 is configured to acquire estimated ex-warehouse information of the target SKU in a future target time period; wherein the estimated ex-warehouse information is determined based on the historical order corresponding to the target SKU;

and the target storage position determining module 705 is configured to determine, according to the estimated inventory information, a target storage position of the target SKU in the target storage area.

Optionally, the encoding module is specifically configured to:

Optionally, the apparatus further includes a word vector model training module, the word vector model training module including:

the order line vector determining unit is used for determining the independent heat expression vector corresponding to each order line in each sample order according to the SKU and the SKU quantity corresponding to each order line in each sample order;

the context vector determining unit is used for determining the context representation vector corresponding to each order line in each sample order according to the independent heat representation vector corresponding to each order line in each sample order;

the input-output pair determining unit is used for determining the independent heat representing vector and the context representing vector corresponding to each order row in the same sample order as an input-output pair;

and the word vector model training unit is used for training the initial word vector model according to the input-output pair to obtain a trained word vector model.

Optionally, the target area determining module includes:

the target group determining unit is used for clustering all SKUs according to the first SKU affine value and a second SKU affine value between every two SKUs in the stored SKUs, and determining a target group corresponding to the target SKU;

a first target area determining unit, configured to determine, if there is an empty storage bit in a storage area corresponding to the target group, a storage area corresponding to the target group as the target storage area; and if the target group is not allocated with a storage area or no empty storage bit exists in the storage area corresponding to the target group, determining the target storage area according to the estimated library outlet information of the target SKU.

Optionally, the target area determining module includes:

a second target area determining unit, configured to determine, if the target SKU is already stored in the warehouse and an empty storage location exists in a storage area corresponding to the target SKU, a storage area corresponding to the target SKU as the target storage area;

and the third target area determining unit is used for determining the target stored SKU with the affine value smaller than the affine threshold value of the first SKLU if the target SKU is not stored in the warehouse or the empty storage position does not exist in the storage area corresponding to the target SKU, and determining the target storage area according to the storage area of the target stored SKU.

Optionally, the third target area determining unit includes:

a target area determining subunit, configured to determine, as the target storage area, a storage area of the target stored SKU that has an empty storage bit and has the smallest affine value of the first SKU if the storage area of the target stored SKU has the empty storage bit; if the storage area of the target stored SKU does not have empty storage bits, determining the target storage area according to the estimated ex-warehouse information of the target SKU.

Optionally, the target area determining module includes:

a region distance determining unit configured to determine, for each candidate region in the warehouse, a distance between the candidate region and each storage region, the candidate region being a storage region in which an empty storage bit exists in the warehouse;

a third SKU affine value determining unit configured to determine a third SKU affine value between the target SKU and a stored SKU in the candidate area according to the first SKU affine value and each of the distances; the third SKU affine value reflects the probability that the target SKU appears in the same order as the stored SKU in the candidate region;

the ex-warehouse index evaluation value determining unit is used for determining an ex-warehouse index evaluation value corresponding to the candidate region according to the estimated ex-warehouse information of the target SKU, the distance between the candidate region and the picking site and the third SKU affine value;

And a fourth target area determining unit configured to determine a candidate area with the smallest evaluation value of the ex-warehouse index as the target storage area.

Optionally, the warehouse is a multi-container concurrent warehouse;

the ex-warehouse index evaluation value determining unit is specifically configured to:

y＝a*popular(sku)*distance(loc)+b*affinity(sku,loc)

Optionally, the warehouse is a self-climbing robot warehouse;

the heuristic value determining unit is specifically configured to:

y＝a*popular(sku)*distance(loc)-b*affinity(sku,loc)

Optionally, the apparatus further includes:

the order vector determining module is used for determining order code vectors of all to-be-placed orders according to third code vectors of all SKUs in all to-be-placed orders;

the order affine value determining module is used for determining an order affine value between every two to-be-ex orders according to the order encoding vectors of the to-be-ex orders, wherein the order affine value reflects the probability that the two to-be-ex orders are picked together;

the order clustering module is used for clustering all orders to be taken out according to the order affine value to obtain a plurality of order classifications;

the wave grouping processing module is used for classifying each order, performing wave grouping processing on each order to be sorted according to the storage position of the SKU in each order to be sorted, and obtaining at least one wave corresponding to the classification, wherein the wave is used for sorting out.

The specific implementation process of the functions corresponding to each module and unit in the device provided in the embodiment of the present application may refer to the method embodiments shown in fig. 1 to 5, and the specific implementation process of the functions corresponding to each module and unit in the device part will not be described herein.

According to the warehouse management device provided by the embodiment, the first coding vector of the target SKU to be warehoused is determined, the first SKU affine value between the target SKU and each stored SKU is determined according to the first coding vector and the second coding vector of each stored SKU in the warehouse, the target storage area of the target SKU is determined according to the first SKU affine value and the storage area of each stored SKU, the estimated warehouse-out information of the target SKU in a future target time period is obtained, the target storage position of the target SKU to be warehoused in the target storage area is determined according to the estimated warehouse-out information of the target SKU, the relation between different SKUs is fully considered through the first SKU affine value when the target storage area of the target SKU is determined, the warehouse-in of the target SKU to be warehoused is guided based on the first SKU affine value, the relation between the target SKU to be warehoused and the stored SKU is established, the warehouse-in efficiency can be improved, and the whole warehouse-out efficiency of the warehouse can be improved.

For the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments for relevant points.

Fig. 7 is a block diagram of an electronic device provided in an embodiment of the present application, where the electronic device 800 may include one or more processors 810 and one or more memories 820 coupled to the processors 810, as shown in fig. 7. Electronic device 800 may also include an input interface 830 and an output interface 840 for communicating with another apparatus or system. Program code executed by processor 810 may be stored in memory 820.

The processor 810 in the electronic device 800 invokes the program code stored in the memory 820 to perform the warehouse management methods in the above-described embodiments.

According to an embodiment of the present application, there is also provided a computer readable storage medium including, but not limited to, a disk memory, a CD-ROM, an optical memory, etc., having stored thereon a computer program which when executed by a processor implements the warehouse management method of the foregoing embodiment.

According to an embodiment of the present application, there is also provided a computer program product comprising a computer program or computer instructions which, when executed by a processor, implement the warehouse management method as described in the above embodiments.

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.

It will be apparent to those skilled in the art that embodiments of the present application may be provided as a method, apparatus, or computer program product. Accordingly, the present embodiments may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

Embodiments of the present application are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present embodiments have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the present application.

Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.

The foregoing has described in detail a warehouse management method, an electronic device and a storage medium provided in the present application, and specific examples are applied herein to illustrate the principles and embodiments of the present application, where the foregoing examples are only used to help understand the method and core idea of the present application; meanwhile, as those skilled in the art will have modifications in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.

Claims

1. A warehouse management method, comprising:

determining a first encoding vector of a target minimum stock keeping unit SKU to be put in storage;

and determining a target storage position of the target SKU in the target storage area according to the estimated ex-warehouse information.

2. The method of claim 1, wherein the determining the first encoding vector for the target SKU to be binned comprises:

3. The method of claim 2, wherein the training process of the word vector model comprises:

determining a single heat expression vector corresponding to each order line in each sample order according to the SKU and the SKU quantity corresponding to each order line in each sample order;

determining a context representation vector corresponding to each order line in each sample order according to the independent heat representation vector corresponding to each order line in each sample order;

determining a single-hot representation vector and a context representation vector corresponding to each order line in the same sample order as an input-output pair;

and training the initial word vector model according to the input-output pair to obtain a trained word vector model.

4. The method of any of claims 1-3, wherein the determining the target holding area for the target SKU based on the first SKU affine value and the storage area for each of the stored SKUs comprises:

clustering all SKUs according to the first SKU affine value and a second SKU affine value between every two SKUs in the stored SKUs, and determining a target group corresponding to the target SKU;

If the storage area corresponding to the target group has empty storage bits, determining the storage area corresponding to the target group as the target storage area;

and if the target group is not allocated with a storage area or no empty storage bit exists in the storage area corresponding to the target group, determining the target storage area according to the estimated library outlet information of the target SKU.

5. The method of any of claims 1-3, wherein the determining the target holding area for the target SKU based on the first SKU affine value and the storage area for each of the stored SKUs comprises:

if the target SKU is stored in the warehouse and the empty storage bit exists in the storage area corresponding to the target SKU, determining the storage area corresponding to the target SKU as the target storage area;

if the target SKU is not stored in the warehouse or no empty storage bit exists in a storage area corresponding to the target SKU, determining the target stored SKU of which the affine value of the first SKU is smaller than an affine threshold value, and determining the target storage area according to the storage area of the target stored SKU.

6. The method of claim 5, wherein said determining said target holding area based on a storage area of said target stored SKU comprises:

If the storage area of the target stored SKU has the empty storage bit, determining the storage area of the target stored SKU which has the empty storage bit and has the minimum affine value of the first SKU as the target storage area;

if the storage area of the target stored SKU does not have empty storage bits, determining the target storage area according to the estimated ex-warehouse information of the target SKU.

7. The method of any of claims 1-3, wherein the determining the target holding area for the target SKU based on the first SKU affine value and the storage area for each of the stored SKUs comprises:

determining a distance between each candidate region and each storage region for each candidate region in the warehouse, wherein the candidate regions are storage regions with empty storage bits in the warehouse;

determining a third SKU affine value between the target SKU and the stored SKUs in the candidate region according to the first SKU affine value and each distance; the third SKU affine value reflects the probability that the target SKU appears in the same order as the stored SKU in the candidate region;

determining a delivery index evaluation value corresponding to the candidate region according to the estimated delivery information of the target SKU, the distance between the candidate region and the picking site and the third SKU affine value;

And determining the candidate area with the minimum evaluation value of the ex-warehouse index as the target storage area.

8. The method of claim 7, wherein the warehouse is a multi-container concurrent warehouse;

y＝a*popular(sku)*distance(loc)+b*affinity(sku,loc)

9. The method of claim 7, wherein the warehouse is a self-climbing robotic warehouse;

y＝a*popular(sku)*distance(loc)-b*affinity(sku,loc)

10. The method as recited in claim 1, further comprising:

determining order code vectors corresponding to all to-be-placed orders according to third code vectors of all SKUs in all to-be-placed orders;

according to the order encoding vector of each order to be picked, determining an order affine value between every two orders to be picked, wherein the order affine value reflects the probability that the two orders to be picked together;

clustering all orders to be taken out according to the affine value of the orders to obtain a plurality of order classifications;

and aiming at each order classification, carrying out wave grouping processing on each order to be delivered in the classification according to the storage position of the SKU in each order to be delivered, and obtaining at least one wave corresponding to the classification, wherein the wave is used for delivering and picking.

11. An electronic device, comprising: a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor implements the warehouse management method as claimed in any one of claims 1-10.

12. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the warehouse management method as claimed in any of claims 1-10.

13. A computer program product comprising a computer program or computer instructions which, when executed by a processor, implement the warehouse management method of any of claims 1 to 10.