Shop fence information dynamic adjustment method, equipment and medium

Abstract

The embodiment of the invention discloses a shop fence information dynamic adjustment method and device and a medium, and the method comprises the steps: obtaining a plurality of original user positions and a plurality of shop positions; correcting the distance between every two original user positions according to the distance between the original user positions and the co-occurrence frequency in the historical distribution behavior of the rider; clustering the plurality of original user positions according to each corrected distance, so that the original user positions with closer distances and / or more co-occurrence times are more easily clustered into one class; taking each clustering center as each clustering user position; operating an integer programming model by taking the maximization of the total order quantity of the plurality of shops as a target, and optimizing the distribution relationship between the positions of the clustering users and the positions of the shops; and generating fence information of each shop according to the optimized distribution relationship. According to the embodiment, the shop fence information can be automatically and dynamically adjusted.

Description

[0001] This application is a divisional application of the invention patent application with application number 202311699606.4, application date December 12, 2023, and invention name "A method, device and medium for dynamically adjusting store fence information." Technical Field

[0002] The embodiments of the present invention relate to the field of data processing technology, and in particular to a method, device and medium for dynamically adjusting store fence information. Background Art

[0003] Store fence information refers to the geographical scope that a store can serve, and is usually displayed on the map as a continuous area centered on the store. For online shopping systems, the system divides users into fence-bound stores based on the latitude and longitude of the user's delivery address. Users can usually only find the items they need from this store, and the riders perform delivery tasks in the designated area according to the orders assigned by the system. Therefore, store fence information plays a vital role in improving user experience and increasing business order volume.

[0004] In the existing technology, business personnel are usually only allowed to bind fences and stores or delete or modify fences. Business personnel mainly consider the store's geographical location, delivery range, surrounding traffic conditions, and user residence, etc. Fences between stores usually do not cross and there is a clear distance between them. This method is suitable for situations with fewer stores and can meet basic user order needs to a certain extent.

[0005] However, the fences divided manually are not accurate enough. There are distances between fences of neighboring stores, which limits users' opportunities to explore nearby stores' products and prevents them from browsing and comparing the product information of neighboring stores, increasing the occurrence of order loss. At the same time, the division method based on expert experience considers fewer information dimensions and is difficult to take into account other information of stores, users, and riders besides their geographic locations. Once the coordinates of the store fence are set, they are usually not adjusted and cannot meet the real-time changing order needs. Summary of the invention

[0006] The embodiments of the present application provide a method, device and medium for dynamically adjusting store fence information to solve the above technical problems.

[0007] In a first aspect, an embodiment of the present invention provides a method for dynamically adjusting store fence information, comprising:

[0008] Get multiple original user locations and multiple store locations;

[0009] According to the distance between the original user locations and the number of co-occurrences in the rider’s historical delivery behavior, the distance between every two original user locations is corrected:

[0010] Clustering the multiple original user locations according to the corrected distances, so that original user locations with closer distances and / or more co-occurrence times are more likely to be clustered into one category;

[0011] The center of each cluster is used as the location of each cluster user;

[0012] An integer programming model is run with the goal of maximizing the total order volume of the plurality of stores, and the distribution relationship between each clustered user location and each store location is optimized;

[0013] Generate fence information for each store based on the optimized distribution relationship;

[0014] Among them, dist ij represents the distance between the original user position i and the original user position j; cnt ij Indicates the number of co-occurrences of original user location i and original user location j during a delivery by the rider; cluster_dist ij represents the corrected distance between the original user position i and the original user position j.

[0015] In a second aspect, an embodiment of the present invention provides an electronic device, the electronic device comprising:

[0016] one or more processors;

[0017] a memory for storing one or more programs,

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the method for dynamically adjusting store fence information described in any embodiment.

[0019] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for dynamically adjusting store fence information described in any embodiment.

[0020] The embodiment of the present invention provides a method for dynamically adjusting store fence information based on multi-task clustering and operations optimization models, which solves the problems of poor exposure effect, poor delivery timeliness, and long adjustment timeliness in manually dividing fences through an automatically executable method. First, the user locations and stores to be planned are obtained, and the user locations are clustered based on the number of location co-occurrences and distance indicators in the historical delivery data to reduce the number of user locations. The remaining user-store matching pairs are used as optimization objects of the integer programming model to optimize whether there is a delivery relationship. The model uses the total order volume as the objective function to obtain the optimal delivery relationship for each matching pair, thereby generating the fence contour coordinates of the store. This method makes full use of the multi-dimensional information of orders, riders, merchants, and users, taking into account multiple needs such as store, user, and rider resource allocation, and processes a small amount of data and has a fast optimization speed. It can dynamically adjust the fence information according to changes in market conditions and user needs, and improve the ability to respond to rapid changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 is a flow chart of a method for dynamically adjusting store fence information provided by an embodiment of the present invention;

[0023] Figure 2 is a flow chart of another method for dynamically adjusting store fence information provided by an embodiment of the present invention;

[0024] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0026] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0027] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0028] Figure 1 1 is a flow chart of a method for dynamically adjusting store fence information provided by an embodiment of the present invention. The method is applicable to the case where store fence information is dynamically updated according to order data and is executed by an electronic device. Figure 1 As shown, the method specifically includes:

[0029] S110. Acquire multiple original user locations and multiple store locations.

[0030] The locations here are all geographical locations. This step obtains multiple user locations and store locations. In subsequent operations, the distribution relationship between these user locations and store locations will be re-planned to generate new store fence information.

[0031] In a specific implementation, first, historical order data in a certain map area is obtained, and multiple user delivery addresses and multiple store locations in the order data are picked up, which can represent hot spots within a period of time and reflect the real-time characteristics of order demand. Then, the map area is divided into grids, and the grid covering any user delivery address is used as the user location. Since there are a lot of user delivery addresses in the massive order data, this step uses grid division to replace all user delivery locations in the grid with a unified grid to reduce the amount of data for subsequent operations. Optionally, Figure 2 is a flowchart of another method for dynamically adjusting store fence information provided by an embodiment of the present invention, combined with Figure 2, the GeoHash algorithm can be used to encode the GPS location in the user's delivery address. This method can divide the map area into grids and encode each grid, converting different geographical locations into GeoHash coordinates. User delivery addresses falling in the same grid have the same GeoHash coordinates and are regarded as the same user location.

[0032] Further, after obtaining multiple user locations and multiple store locations from historical orders, each user location can be combined with each store location to form multiple matching pairs. For example, there are 3000 user locations and 200 store locations, then each user location is combined with each store location to form 3000×200 matching pairs, and the subsequent operation will plan whether each matching pair has a distribution relationship. Then, the distance between the original user location and the store location in each matching pair is calculated, and the matching pairs whose distance exceeds the set threshold are eliminated; all user locations in the remaining matching pair set are used as the final multiple user locations. For example, after eliminating the matching pairs whose distance is less than the threshold from the original 3000×200 matching pairs, 2000×200 matching pairs remain. If the distance is too far, the meaning of distribution is lost. This step eliminates the matching pairs greater than a certain distance, further reducing the user locations to be planned. Among them, the threshold can be determined based on the farthest distance that the rider can deliver, or the farthest delivery distance in the historical order, which is not limited in this embodiment.

[0033] S120. Correct the distance between every two original user locations based on the distance between the original user locations and the number of co-occurrences in the rider's historical delivery behavior.

[0034] This step clusters multiple user locations, and uses the cluster center to replace all user locations in the same cluster. The purpose is still to reduce the user locations to be planned and further reduce the operating burden of subsequent operations. For the convenience of distinction and description, in this embodiment, the user location before clustering is called the original user location, and the user location after clustering is called the clustered user location. It can be understood that the number of clustered user locations is less than or equal to the number of original user locations.

[0035] Clustering can be performed based on the distance between the original user locations and the number of times they co-occur in the rider's historical delivery behavior. Original user locations with closer distances and / or more co-occurrences are more likely to be clustered into one category. Specifically, the distance between each two original user locations is corrected based on the number of times they co-occur in the rider's historical delivery behavior:

[0036]

[0037] Among them, dist ijrepresents the distance between the original user position i and the original user position j; cnt ij Indicates the number of co-occurrences of original user location i and original user location j during a delivery by the rider; cluster_dist ij Indicates the corrected distance between the original user location i and the original user location j, which can be called the clustering distance. This distance characterizes the possibility that every two original user locations have the same distribution relationship and can be used as the basis for clustering. In particular, when the original user location i and the original user location j are both grids, the same grid may cover multiple user GPS locations (assuming there are Q). At this time, the distance to the other Q-1 GPS locations and the smallest GPS location can be selected in the grid as the longitude and latitude position of the grid; then the distance between the grids is calculated through the longitude and latitude positions of the two grids. ij .

[0038] S130: Cluster the multiple original user locations according to the corrected distances, so that original user locations with closer distances and / or more co-occurrence times are more easily clustered into one category.

[0039] After determining the clustering distance, the multiple original user positions are clustered according to the clustering distance, and the user positions that appear more frequently in the same batch of riders and the user positions that are closer in distance are grouped into a large new block as the processing object for subsequent operations.

[0040] S140: Taking each cluster center as the user location of each cluster.

[0041] In a specific embodiment, clustering the multiple original user positions according to the corrected distances includes: traversing the original user positions, and performing the following operations for each traversed original user position: if the original user position does not belong to any clustering cluster, for each existing clustering cluster, determine whether the distance between the original user position and the cluster center of the clustering cluster is less than a preset threshold; if so, the original user position is divided into the clustering cluster; otherwise, a new clustering cluster is added with the original user position as the clustering center. If the original user position belongs to any existing clustering cluster, for other clustering clusters other than the any clustering cluster, determine whether the distance between the original user position and the cluster center of the any clustering cluster is greater than the distance between the original user position and the cluster center of the other clustering clusters; if so, the original user position is divided into the other clustering clusters. The above traversal process can be performed once or multiple times, and the final cluster centers are used as the clustered user positions.

[0042] For example, the 2000×200 matching pairs obtained in S110 include 2000 original user positions, which are clustered to obtain 1500 clustered user positions, and the matching pairs to be planned are reduced to 1500×200. It can be seen that through the above-mentioned grid division, distance constraint and position clustering, the number of matching pairs is gradually reduced from 2000×200 to 1500×200, which greatly reduces the data processing burden of subsequent operations.

[0043] S150, running an integer programming model with the goal of maximizing the total order volume of the multiple stores, and optimizing the distribution relationship between each clustered user location and each store location.

[0044] This step uses numerical form to express the distribution relationship between each cluster user location and each store location, and constructs a reasonable integer programming model to optimize each numerical solution to maximize the total order volume of all stores.

[0045] In one embodiment, S n,m Indicates whether there is a distribution relationship between the nth cluster user location and the mth store location. When S n,m = 0, it means that there is no distribution relationship between the nth cluster user location and the mth store location; when S n,m =1, it means that there is a distribution relationship between the nth cluster user location and the mth store location. For example, the number of cluster user locations is N, the number of store locations is M, and there are a total of N×M S n,m The value of needs to be re-planned, N×M S n,m Each value combination of constitutes a distribution relationship between each cluster user location and each store location. The purpose of this step is to determine the optimal value combination. Optionally, the following objective function can be set for the integer programming model:

[0046]

[0047] Among them, E n represents the order volume of the nth clustered user location; when the clustered user location corresponds to a certain block (or grid), E n Represents the total amount of orders within the nth block (or grid).

[0048] At the same time, the following constraints are constructed for the integer programming model:

[0049] Condition 1.

[0050] Among them, V n represents the set of stores that can provide delivery services for the nth cluster user location, D n,mrepresents the distance between the nth cluster user location and the mth store in the set, T m represents the delivery capability score of the mth store; λ represents the adjustable delivery distance of the nth cluster user location, which can be a fixed value applicable to all cluster user locations, such as 1km, or set for each cluster user location. If the store closest to the nth cluster user location is denoted as v min , then d n and r n Respectively represent the store v min The distance to the nth cluster user location and the delivery capability score of the store. n It is called the minimum delivery distance, which is determined by the minimum delivery distance d of the nth cluster user location. n The sum of the adjustable delivery distance λ limits the delivery range of the clustered user location. The stores that can provide delivery services for the clustered user location must be located within the delivery range.

[0051] In addition, this condition makes another correction to the distance between the user and the store based on the store's delivery capability score. In order to facilitate the distinction, the above D n,m and d n The corrected distance is called the true distance. It is called the virtual distance between the nth cluster user location and the mth store. This virtual distance can take into account the different delivery capabilities of the stores. The stronger the delivery capability, the shorter the delivery time at the same real distance, which is equivalent to shortening the delivery distance; the worse the delivery capability, the longer the delivery time at the same real distance, which is equivalent to increasing the delivery distance. Delivery range n +λ comparison, the store that can provide services to the nth cluster user location must be a store within the delivery range at a virtual distance, which can prevent the store fence from being too large and exceeding the actual delivery range.

[0052] Furthermore, the delivery capability score represents the comprehensive delivery capability of the store, which can be obtained by comprehensively evaluating the on-time delivery rate, delivery distance and delivery time of the store's orders. Optionally, the scoring formula is as follows:

[0053]

[0054] Wherein, score represents the delivery capability score of the current store, OT represents the on-time delivery rate of the current store, Distance_mean represents the average delivery distance of the current store, Time_mean represents the average delivery time of the current store, Order represents the order volume of the current store, Area represents the fence area of ​​the current store (for example, the fence area divided by the method in the background technology), w1 ,w 2 ,w 3 In practical applications, the values ​​of the variables in formula (5) can be obtained by statistics of historical orders or other data channels, and this embodiment does not impose any specific restrictions.

[0055] Condition 2.

[0056] Among them, C n Indicates the number of stores that can provide delivery services for the nth cluster user location. Optionally, the number of stores within the delivery range of the nth cluster user location can be used as the number of stores that can provide delivery services for the cluster user location. This condition can ensure that each cluster user location has a delivery relationship with at least one store.

[0057] Condition 3.

[0058] Among them, O m represents the order volume of the mth store, β represents the hyperparameter of the order volume growth multiple, is the total number of orders of each clustered user location that has a delivery relationship with the mth store, representing the total number of orders of the mth store after re-planning. This condition prohibits the order quantity from exceeding the corresponding multiple of the original order quantity of the store, which can prevent the store's order volume from skyrocketing and exceeding the store's operating capacity.

[0059] In summary, the mathematical programming model can be simplified as follows:

[0060]

[0061] st

[0062]

[0063] S n,m ∈{0,1},n∈{1,2…N},m∈{1,2…M}(7);

[0064] Running this mathematical programming model, we can get N×M S n,m The best value combination of S is 1 n,m The corresponding clustered user locations and store locations are user-store pairs that are suitable for allocating a delivery relationship, and all such user-store pairs together constitute the optimal delivery relationship.

[0065] S160: Generate fence information of each store based on the optimized delivery relationship.

[0066] In this step, for each store, at least one clustered user location that has a delivery relationship with the mth store is selected from the above value combination, that is, S n,m = 1, and the continuous area that can cover these matching pairs is used as the fence area of ​​the store. The store fence areas determined in this embodiment are allowed to overlap, such as Figure 2 As shown, this gives shops, users and riders more choices.

[0067] Optionally, when a clustered user location has a distribution relationship with a store, it is considered that all original user locations that belong to the same cluster cluster as the clustered user location have a distribution relationship with the store; the above expansion is performed on each clustered user location, and then all original user locations in each corresponding cluster cluster have a distribution relationship with the store; then a continuous area is constructed to cover all original user locations with a distribution relationship, and the area is used as the fence area of ​​the store. In particular, when the user location is represented in the form of a grid, the area outline can be generated by Alpha-shape to cover all grids that have a distribution relationship with the store, and the area outline is the fence information of the store.

[0068] Furthermore, after the new store fence information goes online, new order data can be counted and the new fence information can be comprehensively evaluated. Optionally, the evaluation indicators include order volume increase rate, store area coverage, rider delivery punctuality rate and other aspects, so as to fully understand the effect of dynamic adjustment and potential improvement points. Among them, for the store area coverage, the overlapping area between different fences will be regarded as the new area added by the store, that is, it will be regarded as an increase in the store area coverage. It has been verified that the dynamic adjustment of stores in this method can overlap the fence area of ​​neighboring stores, increase the store service coverage, meet the user's regional hot spot needs, and significantly increase the order volume; at the same time, this method balances the store delivery capabilities between neighboring stores based on historical order data, and can optimize the allocation of rider resources; users can also browse the product information of neighboring stores to reduce the loss of orders caused by missing products.

[0069] It can be understood that according to the above steps, the electronic device can be divided into multiple logic modules, including a data preprocessing module, a clustering module, an operations optimization module and an evaluation module, which are respectively used to perform the above steps, such as Figure 2 shown.

[0070] In summary, the present invention provides a method for dynamically adjusting store fence information based on multi-task clustering and operations optimization model, and solves the problems of poor exposure effect, poor delivery timeliness and long adjustment timeliness faced by manual fence division through an automatically executable method. First, this implementation pre-processes massive historical order data, reduces the number of user locations with matching through GeoHash grid division and distance constraints, and removes invalid user-store matching pairs; then, based on the number of delivery co-occurrences and distance indicators in the historical data, multiple GeoHash grids are clustered to further reduce the user locations to be planned, and the remaining user-store matching pairs are used as the optimization objects of the integer programming model to optimize whether there is a delivery relationship. In order to achieve this goal, the integer programming model uses the total order volume as the objective function, and constructs a virtual distance according to the store delivery capacity, restricts the store's fence area and order increment, and assigns each user location to at least one store to obtain the optimal delivery relationship, thereby generating the store's fence outline coordinates. After the new store fence coordinates are online, the new order data is statistically evaluated for a comprehensive understanding of the effect of dynamic adjustment and potential improvement points. In addition, this method makes full use of the multi-dimensional information of orders, riders, merchants, and users, taking into account multiple needs such as store, user, and rider resource allocation. It processes a small amount of data and has a fast optimization speed. It can dynamically adjust fence information according to changes in market conditions and user needs, and improve its ability to respond to rapid changes.

[0071] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown, the device includes a processor 60, a memory 61, an input device 62 and an output device 63; the number of processors 60 in the device can be one or more. Figure 3 A processor 60 is taken as an example; the processor 60, the memory 61, the input device 62 and the output device 63 in the device can be connected by a bus or other means. Figure 3 The example of connecting through bus is taken in the following.

[0072] The memory 61 is a computer-readable storage medium that can be used to store software programs, computer executable programs and modules, such as program instructions / modules corresponding to the store fence information dynamic adjustment method in the embodiment of the present invention. The processor 60 executes various functional applications and data processing of the device by running the software programs, instructions and modules stored in the memory 61, that is, realizing the above-mentioned store fence information dynamic adjustment method.

[0073] The memory 61 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 61 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 61 may further include a memory remotely arranged relative to the processor 60, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0074] The input device 62 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 63 may include a display device such as a display screen.

[0075] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the method for dynamically adjusting store fence information of any embodiment is implemented.

[0076] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document,

[0077] Computer readable storage media may be any tangible media that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device.

[0078] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0079] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0080] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamically adjusting store fence information, characterized in that: include: Get multiple original user locations and multiple store locations; According to the distance between the original user locations and the number of co-occurrences in the rider’s historical delivery behavior, the distance between every two original user locations is corrected: Clustering the multiple original user locations according to the corrected distances, so that original user locations with closer distances and / or more co-occurrence times are more likely to be clustered into one category; The center of each cluster is used as the location of each cluster user; An integer programming model is run with the goal of maximizing the total order volume of the plurality of stores, and the distribution relationship between each clustered user location and each store location is optimized; Generate fence information for each store based on the optimized distribution relationship; Among them, dist ij represents the distance between the original user position i and the original user position j; cnt ij Indicates the number of co-occurrences of original user location i and original user location j during a delivery by the rider; cluster_dist ij represents the corrected distance between the original user position i and the original user position j.

2. The method according to claim 1, characterized in that The obtaining of multiple original user locations and multiple store locations includes: Extract multiple user delivery addresses and multiple store locations from historical orders; The map area is divided into grids, and the grid covering any user's delivery address is taken as the original user location.

3. The method according to claim 1, characterized in that: The obtaining of multiple original user locations and multiple store locations includes: Combine each original user location with each store location to form multiple matching pairs; Calculate the distance between the original user location and the store location in each matching pair, and remove matching pairs whose distance exceeds a set threshold; All the original user positions in the remaining matching pair set are taken as the final multiple original user positions.

4. The method according to claim 1, characterized in that Clustering the multiple original user locations according to the corrected distances includes: Traverse each original user position, and perform the following operations for each traversed original user position: If the original user position does not belong to any cluster, for each existing cluster, determine whether the distance between the original user position and the cluster center of the cluster is less than a preset threshold; if so, the original user position is divided into the cluster; otherwise, a new cluster is added with the original user position as the cluster center; If the original user position belongs to any existing cluster, for other clusters other than the said cluster, determine whether the distance between the original user position and the cluster center of the said any cluster is greater than the distance between the original user position and the cluster center of the said other cluster; if so, the original user position is divided into the said other cluster.

5. The method according to claim 1, characterized in that The step of running the integer programming model with the goal of maximizing the total order volume of the multiple stores to optimize the distribution relationship between each clustered user location and each store location includes: The attribute value of whether there is a distribution relationship between each cluster user location and each store location is used as the optimization object and input into the integer programming model; With the goal of maximizing the order volume of the multiple stores, and with at least one of the following conditions as a constraint, the integer programming model is run to obtain the optimal value combination of each attribute value: Condition 1: All stores serving each clustered user location are located within the delivery range; Condition 2: Each clustered user location has a delivery relationship with at least one store; Condition 3: The total number of orders from each clustered user location that has a delivery relationship with each store does not exceed the store's operating capacity.

6. The method according to claim 5, characterized in that Condition 1 is achieved by: According to the delivery capability of any store, the real distance between the store and any clustered user location is corrected, so that under the same real distance, the store with stronger delivery capability has a smaller corrected distance; Determine whether the corrected distance exceeds the delivery range.

7. The method according to claim 6, characterized in that Before correcting the actual distance between the store and any clustered user location according to the delivery capability of any store, the method further includes: Each store’s delivery capability is scored based on its delivery punctuality rate, delivery distance and delivery time.

8. The method according to claim 1, characterized in that: The fence information of each store is generated according to the optimized distribution relationship, including: Extract multiple clustered user locations that have a delivery relationship with any store; Determining a plurality of original user locations that belong to the same cluster cluster as the plurality of clustered user locations; constructing a continuous area covering the plurality of original user locations as fence information of the store; Among them, the fence areas of each store are allowed to overlap.

9. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method for dynamically adjusting store fence information as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for dynamically adjusting store fence information as described in any one of claims 1-8 is implemented.

Images