Distribution range generation method and device, program product and storage medium

Abstract

The invention provides a distribution range generation method and device, a program product and a storage medium. The method comprises the steps of obtaining a target merchant of which a distribution range is to be generated and a plurality of candidate user AOIs matched with the target merchant; obtaining the distance between each candidate user AOI and the target merchant; taking maximization of the total delivery order scale of the target merchant as an optimization target, and selecting from the candidate user AOIs by an optimization model to solve an optimal user AOI combination meeting the optimization target and constraint conditions; wherein the constraint condition comprises that under the condition that the AOI of the target user is not selected, the AOIs of other farther users are not selected; and determining a delivery range of the target merchant based on the optimal user AOI combination.

Description

Technical Field

[0001] This specification relates to the field of logistics and distribution technology, and in particular to methods, devices, program products, and storage media for generating a distribution range. Background Art

[0002] In the field of logistics and distribution, it's necessary to generate a corresponding logistics and distribution service range for each target. For example, for a merchant, their delivery range represents their maximum logistics and distribution service area. In other words, orders from the merchant can only be served within this delivery range; locations outside this range are unavailable. Therefore, generating an accurate delivery range for merchants has become a pressing technical challenge. Summary of the Invention

[0003] To overcome the problems existing in the related art, this specification provides a method, device, program product and storage medium for generating a delivery range.

[0004] According to a first aspect of an embodiment of this specification, a method for generating a delivery range is provided, the method comprising:

[0005] Obtain a target merchant for which a delivery range is to be generated and multiple candidate user AOIs matching the target merchant;

[0006] Obtain the distance between each candidate user's AOI and the target merchant;

[0007] With maximizing the total delivery order size of the target merchant as the optimization goal, the optimization model selects from each candidate user AOI to solve the optimal user AOI combination that meets the optimization goal and the constraints. The constraints include: if the target user AOI is not selected, other user AOIs that are farther away are not selected.

[0008] Based on the optimal user AOI combination, the delivery range of the target merchant is determined.

[0009] According to a second aspect of the embodiments of this specification, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method embodiment described in the first aspect are implemented.

[0010] According to a third aspect of the embodiments of this specification, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method embodiment described in the first aspect are implemented.

[0011] According to a fourth aspect of the embodiments of this specification, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of the method embodiment described in the first aspect.

[0012] The technical solutions provided by the embodiments of this specification may have the following beneficial effects:

[0013] In an embodiment of this specification, to solve the problem of not allowing hollow areas to appear within the final generated delivery range, this embodiment designs an optimization model to screen and select from all candidate user AOIs. The constraint condition is designed such that if a relatively close target user AOI is not selected, then other user AOIs that are farther away will also not be selected. This ensures the geographic consistency of user AOI selection, avoids skipping over nearby user AOI areas and selecting distant user AOIs, and prevents the appearance of hollow areas in the determined delivery range. In this way, the geographic constraint is approximately incorporated into the optimization model, enabling a rapid and accurate solution. The optimization model can select a user AOI combination that meets the optimization objective. The optimization objective is designed to maximize the total delivery order size of the target merchant. The optimal user AOI combination ultimately solved by the optimization model generates the largest total delivery order size. This embodiment solves the problem of accurate solution in delivery range generation to a certain extent, making range generation more reasonable and accurate.

[0014] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of a delivery scenario shown in this specification according to an exemplary embodiment.

[0016] Figure 2A This is a flowchart of a method for generating a delivery range according to an exemplary embodiment of this specification.

[0017] Figure 2B This is a schematic diagram of a segmentation method according to an exemplary embodiment of this specification.

[0018] Figure 2C FIG. 1 is a schematic diagram of a candidate user AOI of a target merchant according to an exemplary embodiment of the present specification.

[0019] Figure 3 This is a hardware structure diagram of a computer device where a device for generating a delivery range is located according to an exemplary embodiment of the present specification.

[0020] Figure 4This is a block diagram of a device for generating a delivery range according to an exemplary embodiment of the present specification. DETAILED DESCRIPTION

[0021] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with this specification. Rather, they are merely examples of apparatus and methods consistent with certain aspects of this specification, as detailed in the appended claims.

[0022] The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit this specification. As used in this specification and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0023] It should be understood that although the terms first, second, third, etc. may be used in this specification to describe various information, such information should not be limited to these terms. These terms are merely used to distinguish information of the same type from one another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information without departing from the scope of this specification. Depending on the context, the term "if" as used herein may be interpreted as "when," "when," or "in response to determining."

[0024] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this manual are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0025] In the field of logistics and distribution, the delivery range is one of the key factors in the merchant's purchasing service standards. It directly affects the merchant's traffic limit and is of great significance to the achievement of logistics service standards.

[0026] Existing scope generation is mostly based on geofence information: given the merchant's latitude and longitude points, the corresponding radius limit, and the service product's requirements for fulfillment efficiency and quality, the corresponding user interest area (AOI) is screened to approximately achieve the corresponding goal.

[0027] A typical strategy is to use heuristic methods to generate closed polygons by repeatedly adding or removing AOIs. The main problem with this approach is that the resulting range is not necessarily the optimal solution, and there is a certain degree of error, which can potentially impact business operations. The main reason for not using a direct and precise solution is that it involves modeling a geographic location constraint: the resulting polygon must be closed and cannot have any hollow areas in the middle, a constraint that is difficult to directly reflect during modeling.

[0028] In the embodiments of this specification, by adding geographic constraints to the optimization model, an accurate solution can be quickly obtained, which solves the problem of accurate solution in range generation to a certain extent, making range generation more reasonable and accurate.

[0029] like Figure 1 Shown is a schematic diagram of a delivery scenario according to an exemplary embodiment of this specification, including a service party, delivery capacity and users. Among them, the service party has built a service end and provided a client to the user, and the user can use the services provided by the business party through the client; different from the client for users, the business party also provides the store party with a client for the store party, and the store party can use the service provided by the business party through the client. Among them, users can view the stores that can provide services for them through the client, can trade with the store party, and can initiate delivery orders; the business party can allocate delivery capacity for the instant delivery order. The delivery capacity of this embodiment refers to the party with delivery capabilities, including but not limited to delivery personnel, such as the so-called riders. The delivery capacity can communicate with the service end through the client used for the delivery capacity; in other examples, the delivery capacity can also include unmanned delivery equipment, such as drones, unmanned vehicles, etc.

[0030] Service providers need to determine the merchant's delivery range. For example, for a merchant, the merchant's delivery range can represent the maximum logistics and delivery service area within the business district. Once the merchant's delivery range is determined, it can be used as a reference for online logistics and delivery services within the business district. For example, merchants can be exposed to users within the merchant's delivery range. In other words, the merchant can provide services to users within the merchant's delivery range, but may not provide services to users outside the merchant's delivery range.

[0031] For example, after defining a merchant's delivery range, if a user needs to use an online service, if the user is within the merchant's delivery range, the user can discover and use the merchant's online services, such as food delivery services, on the client. However, if the user is outside the merchant's delivery range, the user may not be able to discover and use the merchant's online services on the client.

[0032] like Figure 2A As shown, Figure 2A This is a flowchart of a method for generating a delivery range according to an exemplary embodiment of the present specification, including the following steps:

[0033] In step 202, a target merchant for which a delivery range is to be generated and a plurality of candidate user AOIs matching the target merchant are obtained.

[0034] In step 204, the distance between each candidate user AOI and the target merchant is obtained.

[0035] In step 206 , with maximizing the total delivery order size of the target merchant as the optimization goal, the optimization model selects from the candidate user AOIs to solve the optimal user AOI combination that meets the optimization goal and the constraints.

[0036] The constraint condition includes: when the target user AOI is not selected, other user AOIs that are farther away are not selected.

[0037] In step 208, based on the optimal user AOI combination, the delivery range of the target merchant is determined.

[0038] The method for generating a delivery range in this embodiment can be run on a computer device, including but not limited to a server, a cloud server, a server cluster, a tablet computer, a personal digital assistant, a laptop computer, or a desktop computer. The method in this embodiment can be applied to various application scenarios, such as Figure 1 In the application scenario shown, the service provider can apply the solution of this embodiment to generate the delivery range of the target merchant in advance or in real time.

[0039] For example, in a geographic information system (GIS), an AOI can be used to represent regional geographic entities on a map, such as residential areas, schools, and scenic spots. A point of interest (POI) can be a house, a shop, a mailbox, a bus stop, and so on. When defining an AOI, the boundary point information can be clearly defined. From a spatial geometry perspective, an AOI is a collection of multiple POI data. Therefore, compared to POIs, AOIs are more expressive and can intuitively convey information such as the data's distribution, area, and density.

[0040] As an example, the delivery service provider may pre-build a map database, which may pre-store geographic data related to one or more user AOIs. Based on the geographic location of the target merchant, all candidate user AOIs of the target merchant may be obtained from the map database.

[0041] As an example, data corresponding to each user's AOI can be prepared in advance, including but not limited to: geographic location range (which can be composed of multiple geographic location coordinates), one or more order information of historical orders (including but not limited to order quantity, order amount, total historical transaction amount, delivery time or receiving address, etc.).

[0042] The multiple candidate user AOIs matching the target merchant in step 202 can be obtained in a variety of ways, which are not limited in this embodiment. For example, these multiple candidate user AOIs can be obtained using relatively loose conditions to obtain an initial but inaccurate delivery range. The specific conditions can be flexibly configured according to actual needs. For example, the range can be obtained based solely on the size of the range, such as using a larger distance threshold with the target merchant's geographic location as the center. For example, all user AOIs within a certain shape (such as a circle, polygon, rectangle, or irregular shape) within a preset distance threshold, centered on the target merchant's geographic location, can be used as candidate user AOIs.

[0043] Among them, the setting of the preset distance threshold can be set according to the actual application scenario; for example, in some application scenarios, it can be determined based on the delivery needs of the target merchant. For example, the delivery service provider can provide different types of delivery services for different merchants, including but not limited to fast delivery services based on short distances, or large-scale delivery services based on long distances. For example, catering merchants are more sensitive to delivery time and have higher timeliness requirements, so they can adopt fast delivery services based on short distances; retail chain merchants often require larger exposure traffic and have lower timeliness requirements, so they can adopt large-scale delivery services based on long distances. Therefore, the preset distance threshold can be determined according to the type of delivery service required by the target merchant; if the target merchant requires a fast service based on a short distance, a smaller preset distance threshold can be used; if the target merchant requires a large-scale delivery service based on a long distance, a larger preset distance threshold can be used. The specific numerical value is not limited in this embodiment.

[0044] Alternatively, the determination can be made in conjunction with other factors, such as the region where the merchant is located. For example, the terrain of the region, for example, is relatively simpler than that of mountainous regions. Therefore, the delivery difficulty in plain regions is less than that in mountainous regions. Therefore, the preset distance threshold in plain regions can be lower than that in plain regions. Alternatively, some regions may have more bridges due to the presence of rivers, which also increases the difficulty of delivery. Therefore, the preset distance threshold in these regions can be lower than that in regions with fewer bridges. Of course, in other examples, obtaining information based on distance thresholds and / or other considerations is also optional, such as combining historical order information of each user's AOI, etc., and this embodiment does not limit this.

[0045] In this embodiment, the multiple candidate user AOIs that match the target merchant may correspond to a relatively broad initial delivery range. Then, a precise delivery range may be solved from these multiple candidate user AOIs. It can be understood that the delivery range finally determined in step 204 is within the initial delivery range. The process of generating a precise delivery range is to determine which user AOIs can be selected and which user AOIs cannot be selected from these multiple candidate user AOIs to meet the optimization goal.

[0046] In this embodiment, to rapidly obtain an accurate solution by approximating geographic constraints into the optimization model, the distance between each candidate user AOI and the target merchant is obtained in step 204. Furthermore, in step 206, a constraint condition is set: if the target user AOI is not selected, then other user AOIs at greater distances are also not selected. In this way, mathematical modeling of geographic constraints is achieved through the distance between the user AOI and the target merchant.

[0047] Among them, through the optimization model, all candidate user AOIs can be screened and selected so that the selected user AOI combination meets the optimization goal; among them, the optimization goal is designed to make the total delivery order scale of the target merchant, then the optimization model can solve the user AOI combination with the largest total delivery order scale of the target merchant, and finally the optimal user AOI combination solved will generate the largest total delivery order scale.

[0048] Among them, in the constraint conditions, if a relatively close target user AOI is not selected, then other user AOIs that are farther away will not be selected. In this way, the geographical consistency of user AOI selection can be guaranteed, avoiding the situation of skipping the nearby user AOI area and selecting the distant user AOI, and preventing the occurrence of hollow areas in the determined delivery range.

[0049] As an example, the total delivery order scale of the target merchant may be the total delivery order quantity, or the total delivery order amount, or other custom parameters representing the delivery order scale, which is not limited in this embodiment.

[0050] In step 208, the optimization model selects multiple user AOIs that meet the optimization objectives and constraints. These user AOIs selected by the optimization model are called the optimal user AOI combination in this embodiment. The geographical area covered by the optimal user AOI combination is the delivery range of the target merchant.

[0051] As can be seen from the above embodiment, in order to solve the problem of no hollow areas appearing within the final generated delivery range, this embodiment designs an optimization model to screen and select from all candidate user AOIs, wherein the constraint condition is designed such that if a relatively close target user AOI is not selected, then other user AOIs that are farther away will not be selected either. In this way, the geographical consistency of the user AOI selection can be guaranteed, and the situation of skipping the nearby user AOI area and selecting the distant user AOI can be avoided, thereby preventing the appearance of hollow areas in the determined delivery range. In this way, the geographic constraint is approximately added to the optimization model, and an accurate solution can be quickly obtained. The optimization model can select a user AOI combination that meets the optimization goal. The optimization goal is designed to maximize the total delivery order scale of the target merchant. The optimal user AOI combination ultimately solved by the optimization model generates the largest total delivery order scale. This embodiment solves the problem of accurate solution in the generation of the delivery range to a certain extent, making the range generation more reasonable and accurate.

[0052] In some examples, the multiple candidate user AOIs matched by the target merchant may be obtained by:

[0053] Taking the geographical location of the target merchant as the center, the geographical range within the preset distance threshold is divided into multiple planes;

[0054] The candidate user AOIs in each plane are acquired in parallel to obtain multiple candidate user AOIs matching the target merchant.

[0055] As an example, a preset geographical range can be obtained with the geographical location of the target merchant as the center. The size and shape of the preset geographical range can be set according to actual needs, and can be a square, rectangle, circle or irregular shape, etc. This embodiment does not limit this.

[0056] The division method may be an average division method, etc., which is not limited in this embodiment. For example, the geographic range may be divided into K planes at equal angles (e.g., 360 / K degrees), and the value of K may be set as needed. That is, the number of divisions may be flexibly set according to actual needs, which is not limited in this embodiment.

[0057] Figure 2B This is a schematic diagram of a segmentation method according to an exemplary embodiment of this specification. For the convenience of illustration, Figure 2B In the example, the geographical location of the target merchant is taken as the center (as shown by the solid circle in the figure), and 8 planes are divided from the center (i.e., ① to ⑧ shown in the figure). The candidate user AOIs in each plane can be obtained respectively (as shown by the hollow circles in the figure); the specific acquisition method is the same as the above embodiment. The preset distance threshold can be used as the acquisition basis, and the user AOIs located in each plane and whose distance to the target merchant is less than or equal to the preset distance threshold are recalled from the database.

[0058] As an example, the user AOI has a certain size, so whether the user AOI is located in a plane can be determined by whether the center point of the user AOI is located within the range of the plane. In actual applications, other methods can also be used as needed, and this embodiment does not limit this.

[0059] As an example, the distance between the user AOI and the target merchant may refer to the distance between the center point of the user AOI and the geographical location of the target merchant. In actual applications, other methods may be used as needed, and this embodiment does not limit this.

[0060] In practice, if the preset distance threshold is long, a large number of candidate user AOIs need to be recalled, and the recall time is long. Optionally, for each plane, a parallel recall method can be used, or a method of parallel recalling for some planes can be used to improve the recall speed.

[0061] In some examples, the multiple candidate user AOIs that match the target merchant may be obtained by:

[0062] Divide the preset geographical range corresponding to the target merchant into multiple planes, and obtain the candidate user AOI in each plane to obtain multiple candidate user AOIs matching the target merchant;

[0063] The constraint condition includes: when the target user AOI is not selected, other user AOIs that are farther away in the plane to which the target user AOI belongs are not selected.

[0064] The method for dividing the preset geographic range corresponding to the target merchant in this embodiment can refer to the above embodiment. In this embodiment, by dividing the preset geographic range corresponding to the target merchant into multiple planes, the constraint conditions can be further refined. Specifically, the constraint conditions can be further refined so that if the target user AOI is not selected, other user AOIs that are farther away within the plane to which the target user AOI belongs are not selected.

[0065] like Figure 2B As shown, assuming that the AOI closest to the target merchant in plane ⑥ is not selected (i.e., the hollow circle marked with an X in the figure), but the AOI farther away in the plane is selected, the generated delivery range will include an AOI that is not included in the service range (i.e., the hollow circle marked with an X in the figure).

[0066] Based on this, this embodiment has designed the aforementioned constraints to prevent this problem. For example, within the same plane, candidate user AOIs can be sorted from closest to farthest from the target merchant. If the sth AOI is not selected, the s+1th and subsequent AOIs further away within the same plane are prohibited from selection. Based on this, the constraints within each plane can be executed independently or in parallel as needed, thereby ensuring continuity and closure in the selection of AOIs within each plane, and ultimately ensuring continuity and closure in the optimal user AOI combination.

[0067] In some examples, the method may further include:

[0068] For each candidate user AOI, a preset prediction model is used to predict the delivery order index between the target merchant and the candidate user AOI;

[0069] The constraint condition also includes: the predicted delivery order index is within a preset index range.

[0070] As an example, there are multiple delivery order indicators, and a corresponding preset indicator range can be set for each delivery order indicator; that is, the constraint condition can include: each delivery order indicator is within the corresponding preset indicator range.

[0071] In actual applications, the delivery order indicators can be one or more, and the specific indicators can be flexibly configured according to actual needs; as examples, they include but are not limited to:

[0072] The number of delivery orders, such as the number of orders that the AOI may generate in the future time period;

[0073] Delivery time, such as the average delivery time from the time a user places an order to the time the user receives the package within the AOI;

[0074] Delivery distance, such as the average distance of the delivery route from the target merchant to the AOI;

[0075] Delivery efficiency is represented by the average number of combined delivery orders of the target merchant and other delivery orders within the AOI. For example, 0 indicates that the delivery order is currently directly assigned to the delivery capacity, and the delivery capacity has no other back orders. 1 indicates that when the delivery capacity is assigned to the delivery order, the delivery capacity has 1 back order, that is, the delivery order can be combined with 1 other delivery order. The same applies to other values. Of course, other numerical representations can also be used in actual applications, and this embodiment does not limit this.

[0076] In actual applications, other indicators may also be used, which are not limited in this embodiment.

[0077] As an example, this embodiment further designs constraints based on the delivery order indicators, that is, each predicted delivery order indicator must meet the preset indicator range corresponding to the indicator; in actual applications, the preset indicator range corresponding to each indicator can be flexibly configured according to actual needs, and this embodiment does not limit this.

[0078] The following shows a specific example of delivery order indicators and the corresponding preset indicator ranges:

[0079] ① The number of delivery orders, such as the number of orders from the target merchant in the future period (such as a custom period of 7 days) within the candidate user's AOI. The corresponding preset indicator range can be an order volume greater than or equal to 100 orders (or other custom values);

[0080] ② Delivery time, such as the average delivery time from the merchant to the candidate user's AOI; the corresponding preset indicator range can be a delivery time less than or equal to 40 minutes (or other custom values);

[0081] ③ Delivery distance, such as the average distance of the delivery path from the merchant to the candidate user AOI. The corresponding preset indicator range can be a delivery distance less than or equal to 5 kilometers (or other custom values).

[0082] As an example, if there are multiple delivery order indicators, each delivery order indicator can be predicted separately by a corresponding prediction model, or one prediction model can predict at least two delivery order indicators.

[0083] For example, the preset prediction model can be a multi-task prediction model that can simultaneously learn multiple tasks, such as a mixture of experts (MMoE). Using a single preset prediction model to predict all delivery order metrics avoids building and training individual models for each, reducing computing resources and training time. Furthermore, multi-task learning allows the model to share information across different metrics, exploring potential connections between them and improving prediction accuracy.

[0084] Next, we will take the multi-task prediction model as an example to illustrate the model training process.

[0085] First, data preparation and feature engineering can be performed. Training data can include but is not limited to the following data:

[0086] The historical order data of each user in the user AOI may include one or more of the following information: order delivery time, order delivery distance, order timestamp (which can be used to distinguish different time periods), information indicating whether the order was delivered on time, etc.

[0087] Merchant data may include one or more of the following information: category, rating, delivery type (such as the short-distance delivery service or long-distance delivery service mentioned in the above example), number of historical orders (which can be broken down to the number of historical orders in different time periods), geographic location, business district attributes, and other information.

[0088] The basic attribute data of user AOI includes but is not limited to the following information: number of users, consumption level, historical order volume, delivery time, delivery distance, road network complexity, delivery difficulty, etc.

[0089] Based on the above data, features can be constructed; for example, numerical feature standardization (such as delivery distance normalization, delivery time normalization, etc.) can be performed, and categorical features can be embedded (such as encoding merchant categories into low-dimensional vectors) and other processing can be performed. This embodiment does not limit this.

[0090] The model architecture can also be designed. For example, an existing multi-task model, such as a hybrid expert model, can be selected; or an existing model architecture can be modified as needed, or a self-designed model architecture can be designed. This embodiment does not limit this.

[0091] Taking the architecture of the hybrid expert model as an example, the hybrid expert model of this embodiment can include an expert network, which can be designed to include multiple independent neural networks, and each expert learns a different feature combination. It can also include a gating network, the number of gating networks is the same as the number of prediction tasks, and each prediction task (i.e., the prediction task of each delivery order indicator) is assigned an independent gating network; the gating network can output weights to determine the contribution ratio of each expert network to the current prediction task. The model can also include a task output layer. After each task is output by the weighted expert, it is connected to an independent fully connected layer to generate the predicted value of each task.

[0092] The loss function of the model can be designed based on each prediction task. For example, a corresponding loss function is designed for each delivery order indicator. The total loss of the model can be the sum of the losses of each delivery order indicator. Alternatively, the total loss of the model can be set as the weighted sum of the losses of each delivery order indicator as needed. The specific weights can be set according to actual needs, and this embodiment does not limit this.

[0093] Based on this, the desired prediction model can be trained using the above method. Thus, when generating a delivery range for a target merchant, the target merchant's merchant data and relevant data about the candidate user AOI (such as the aforementioned historical order data and basic attribute data) can be input into the model, which then predicts the delivery order metrics between the target merchant and the candidate user AOI.

[0094] In actual applications, there may be cases where the target merchant does not have complete merchant data. For example, the merchant data required for model prediction contains multiple pieces of information, but the target merchant's merchant data does not contain all of this information, and some information may be missing. For example, the target merchant may have been with the service provider for a short time and has no historical orders, so the merchant data may be missing information such as the number of historical orders.

[0095] Based on this, in this embodiment, the training data of the preset prediction model includes: merchant data samples of merchants, and the merchant data samples include one or more types of merchant information samples;

[0096] The preset prediction model predicts the delivery order indicator based on the merchant data of the target merchant; the merchant data of the target merchant is obtained in the following manner:

[0097] If the target merchant does not have merchant information of the target type, merchant information of the target type of the target merchant is generated based on merchant information of the target type of similar merchants similar to the target merchant; wherein the similar merchants and the target merchant are located in the same business district and provide the same category of goods.

[0098] In this embodiment, the training data of the preset prediction model may include merchant data. The preset prediction model needs to rely on the merchant data when predicting the delivery order indicators. The merchant data may include one or more different types of merchant information. As an example, it is assumed that the following types of merchant information are included:

[0099] Basic merchant attribute information, such as category (e.g., catering, retail), geographic location (latitude and longitude), service type (delivery / express delivery), business hours, and historical ratings;

[0100] Order-related information, such as historical order volume, average delivery time, on-time delivery rate, and user repurchase rate;

[0101] Business district characteristic information: population density, consumption level index, number of competing businesses, transportation convenience score, etc.

[0102] For target merchants who have just joined the service provider, some types of merchant information may not be stored and cannot be obtained yet. In this case, this embodiment can also generate the merchant information of the target merchant of the target type based on the merchant information of similar merchants of the target type to complete the missing merchant information of the target merchant.

[0103] Among them, similar merchants are designed to have business district consistency, that is, they are located in the same geographical business district as the target merchant. The standard of whether they are in the same business district can be configured according to actual needs, for example, the geographical distance between the two merchants is within a preset range; or, the service provider has pre-divided the scope of each business district, and the determination is based on whether the two merchants are located within the scope of the pre-defined business district.

[0104] Similar merchants also have category consistency with the target merchant and provide the same or highly similar goods / services (such as both merchants are hot pot restaurants, or both are Cantonese restaurants). The specific category can be pre-set by the service provider, and this embodiment does not limit this.

[0105] Optionally, other criteria may be combined to determine similar merchants, for example, factors such as the size of the merchant, etc. The determined similar merchants may be one or more, which is not limited in this embodiment.

[0106] For example, if a target merchant is missing merchant information of the target type, the merchant information of the target merchant can be directly used. If a target merchant is missing merchant information of the target type, the merchant information of the target merchant can be generated by weighting the merchant information of the target type of each similar merchant. For example, the average of the merchant information of the target merchant can be taken, or the similarity between each similar merchant and the target merchant can be used as the weight to obtain the merchant information of the target type.

[0107] In this way, in this embodiment, when the target merchant is a new merchant and has missing merchant information, the missing information can also be supplemented based on the merchant information of similar merchants, solving the model prediction problem in the scenario of missing merchant data.

[0108] In some examples, the delivery order indicator includes: the number of delivery orders; the optimal optimization model includes a mixed integer programming model;

[0109] The decision variable of the mixed integer programming model includes a binary variable for representing whether to select the candidate user AOI; wherein the value of the binary variable representing the selection of the candidate user AOI is greater than the value of the binary variable representing the non-selection of the candidate user AOI;

[0110] The objective function of the mixed integer programming model includes: maximizing the total delivery order size of the target merchant, where the total delivery order size refers to the sum of the products of the binary variables of each candidate user AOI in multiple candidate user AOIs and the number of delivery orders;

[0111] The constraints of the mixed integer programming model include: for multiple candidate user AOIs sorted from near to far based on their distance from the target merchant, if the value of the binary variable of the s-th user AOI indicates that the user AOI was not selected, then the binary variables of the s+1-th and all subsequent user AOIs with a farther distance are all set to the value indicating that the user AOI was not selected.

[0112] In this embodiment, a mixed integer programming model MIP is designed to achieve an accurate solution. Mixed integer programming is a mathematical optimization method used to find the optimal decision under constraints. The characteristic of this model is that, for variable types, some variables are allowed to be integers (such as 0 or 1) and some are continuous variables (such as delivery time and distance). In this embodiment, the optimization model needs to select among the candidate user AOIs, so the decision variables of the mixed integer programming model are designed to include binary variables, and the values ​​of the binary variables are used to represent whether the candidate user AOI is selected. In the delivery range generation scheme of this embodiment, the MIP model selects the optimal combination from multiple candidate AOIs through mathematical modeling while satisfying the set constraints.

[0113] Among them, the binary variable can be defined as: if the candidate user AOI is selected, the corresponding variable x s =1; if not selected, then x s =0.

[0114] In the constraint condition, user AOIs can be sorted from near to far by distance. For example, taking N candidate user AOIs as an example, s = 1, 2, ..., N; if an AOI is not selected (x_s = 0), all AOIs farther than it must meet x s+1 =0,x s+2 =0,...,x N =0. The corresponding mathematical formula can be: s+1 ≤x s ,

[0115] Based on this, the optimization goal can be:

[0116]

[0117] Among them, O s Indicates the delivery order size of the sth user AOI; only the selected user AOI (x s =1) will contribute to the order volume.

[0118] Therefore, this embodiment ensures that the selected user AOI can continuously cover the area extending outward from the merchant location through progressive constraints, avoiding the presence of unselected user AOIs in the middle; and the use of a mixed integer programming model can directly solve the global optimal solution rather than the local optimal solution of the heuristic strategy.

[0119] As mentioned in the above embodiments, in actual applications, all candidate user AOIs can be obtained based on the preset geographic range corresponding to the target merchant and subsequent processes can be executed. Alternatively, the preset geographic range corresponding to the target merchant can be divided into multiple planes, and subsequent processes can be executed based on the candidate user AOIs within each plane. Based on this, in order to apply these two methods, please see the following formula:

[0120]

[0121] Among them, x ks , indicating whether the sth AOI of the kth plane is selected;

[0122] O ks Indicates the user's AOI ks The number of delivery orders;

[0123] S[k] indicates that the entire plane is divided into K planes. For example, if the plane is not divided, K = 1. If the plane is divided, the value of K can be set according to the number of planes, such as K = 8.

[0124] t ks d ks 、h ks , respectively represent the predicted delivery time, delivery distance and delivery efficiency of the user's AOI.

[0125] Specifically, x ks These are the decision variables in the mixed integer programming model; for example, x ks =1 means select the user AOI; x ks =0 indicates that the user AOI is not selected. Of course, other values ​​can also be used in actual applications, and this embodiment does not limit this.

[0126] The optimization goal of the mixed integer programming model is to maximize the total delivery order size of the target merchant, which is the following formula:

[0127]

[0128] This enables the model to select the user AOI combination that can bring the most orders.

[0129] The constraints of the model are as follows:

[0130] The constraints on delivery time are:

[0131]

[0132] This means that the average delivery time of the currently selected user AOI combination does not exceed the preset threshold

[0133] The constraints on delivery distance are:

[0134]

[0135] This means that the average delivery distance of the currently selected user AOI combination does not exceed the preset threshold

[0136] The constraints on distribution efficiency are:

[0137]

[0138] This means that the average delivery efficiency of the currently selected user AOI combination does not exceed the preset threshold

[0139] This represents the closure constraint that when the target user's AOI is not selected, the AOIs of other users farther away are not selected:

[0140] x k(s+1) ≤x ks ,

[0141] Its meaning is: in the same plane k, if the sth AOI is not selected (x ks=0), the s+1th AOI further away cannot be selected (x k(s+1) =0). This ensures that the generated delivery range is a continuous closed area without any internal hollowing.

[0142] In practical applications, an optimization solver can be used for efficient calculations, ultimately obtaining the optimal user AOI combination with the largest total delivery order size.

[0143] Based on this, this embodiment adopts the MIP model, which can directly embed the complex logic such as the above-mentioned closure constraints into the model through mathematical formulas, thereby achieving a direct and accurate solution to the global optimal solution rather than an approximate solution of the heuristic method.

[0144] like Figure 2C As shown, this specification shows a schematic diagram of candidate user AOIs of a target merchant according to an exemplary embodiment. The center point in the figure is the location of the target Shanghai, the other points in the figure are candidate user AOIs, the user AOIs not selected by the optimization model are the user AOIs represented by the deleted red points shown in the figure, and the remaining user AOIs are the optimal user AOI combination.

[0145] Through the above embodiment, after solving for the optimal user AOI combination, we can obtain the user AOI combination that maximizes the total delivery order size, and use this combination to generate a delivery range. The geographic location range covered by all user AOIs in the optimal user AOI combination, i.e., the range within which the target merchant can serve, is also known as the target merchant's delivery range. In actual applications, the specific method for determining the target merchant's delivery range can be configured as needed and is not limited in this embodiment.

[0146] As an example, determining the delivery range of the target merchant based on the optimal user AOI combination may include:

[0147] The delivery range of the target merchant is obtained based on the geographical boundaries of the outermost user AOIs in the optimal user AOI combination that are away from the target merchant.

[0148] For example, the boundaries of the user AOIs farthest from the target merchant are the outermost user AOIs. Connecting these boundaries forms a closed polygon, which can be used to define the target merchant's delivery range. The interior of this closed polygon represents the target merchant's delivery range. This allows the target merchant's delivery range to be quickly and accurately determined based on the optimal user AOI combination.

[0149] In some examples, the method further comprises:

[0150] In response to a store display request from a user located within the delivery range of the target business district, the target merchant is pushed to the user.

[0151] In actual applications, the service provider can use the solution of this embodiment to generate a delivery range for each merchant and store the delivery range of each merchant. In this way, when a user uses the client, the service provider can determine whether the user is within the delivery range of each merchant based on the user's geographical location. For example, in response to a store display request from a user located within the delivery range of the target merchant, the service provider can push the target merchant's store to the user, thereby exposing the target merchant's online store that can provide services to the user.

[0152] For the same merchant, after obtaining its delivery range using the method of this embodiment, considering that the user AOI may change, or the delivery order indicators may change, or the merchant's demand for the total delivery order size may change, the method of this embodiment can also be executed again in the future to generate a new delivery range for the merchant.

[0153] Optionally, after generating a target merchant's delivery range, services can be provided to the target merchant based on the delivery range. Subsequently, based on actual delivery orders, the target merchant's actual delivery order indicators or actual total delivery order size can be continuously obtained. Constraints and / or optimization targets can be adjusted based on actual conditions, and the solution of this embodiment can be executed again to generate a new delivery range for the merchant, and services can be provided to the target merchant based on the new delivery range.

[0154] Corresponding to the aforementioned embodiment of the method for generating a delivery range, this specification also provides an embodiment of an apparatus for generating a delivery range and a computer device used therein.

[0155] The embodiments of the generation device within the scope of distribution of this specification can be applied to computer devices, such as servers or terminal devices. The device embodiments can be implemented through software, hardware, or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by the processor in which it is located reading the corresponding computer program instructions in the non-volatile memory into the memory and running them. From the hardware level, if Figure 3 The figure is a hardware structure diagram of the computer equipment where the generating device of this manual is located, except Figure 3 In addition to the processor, network interface, memory, and non-volatile memory shown, the computer device where the distribution range generating device is located in the embodiment can generally include other hardware according to the actual function of the computer device, which will not be described in detail.

[0156] like Figure 4 As shown, Figure 41 is a block diagram of a device for generating a delivery range according to an exemplary embodiment of the present specification, the device comprising:

[0157] The first acquisition module 41 is used to: acquire a target merchant for which a delivery range is to be generated and a plurality of candidate user AOIs matching the target merchant;

[0158] The second acquisition module 42 is used to obtain the distance between each candidate user AOI and the target merchant;

[0159] Solving module 43 is configured to: maximize the total delivery order volume of the target merchant as an optimization goal, and select from among the candidate user AOIs using an optimization model to solve for an optimal user AOI combination that meets the optimization goal and constraints; wherein the constraints include: if the target user AOI is not selected, then other user AOIs that are farther away are not selected;

[0160] The determination module 44 is configured to determine the delivery range of the target merchant based on the optimal user AOI combination.

[0161] In some examples, the multiple candidate user AOIs that match the target merchant are obtained by:

[0162] Divide the preset geographical range corresponding to the target merchant into multiple planes, and obtain the candidate user AOI in each plane to obtain multiple candidate user AOIs matching the target merchant;

[0163] The constraint condition includes: when the target user AOI is not selected, other user AOIs that are farther away in the plane to which the target user AOI belongs are not selected.

[0164] In some examples, the delivery order indicator includes: the number of delivery orders; the optimal optimization model includes a mixed integer programming model;

[0165] The decision variables of the mixed integer programming model include binary variables for representing whether to select the candidate user AOI;

[0166] The objective function of the mixed integer programming model includes: maximizing the total delivery order size of the target merchant, where the total delivery order size refers to the sum of the products of the binary variables of each candidate user AOI in multiple candidate user AOIs and the number of delivery orders;

[0167] The constraints of the mixed integer programming model include: for multiple candidate user AOIs sorted from near to far based on their distance from the target merchant, if the value of the binary variable of the s-th user AOI indicates that the user AOI was not selected, then the binary variables of the s+1-th and all subsequent user AOIs with a farther distance are all set to the value indicating that the user AOI was not selected.

[0168] In some examples, the apparatus further includes a prediction module configured to: for each candidate user AOI, predict a delivery order indicator between the target merchant and the candidate user AOI using a preset prediction model;

[0169] The constraint condition also includes: the predicted delivery order index is within a preset index range.

[0170] In some examples, the delivery order indicators include: number of delivery orders, delivery time, delivery distance, and delivery efficiency;

[0171] The prediction model includes a hybrid expert model, which is used to predict each of the delivery order indicators based on the input merchant data of the target merchant, the historical order data and basic attribute data of the candidate user AOI.

[0172] In some examples, the training data of the preset prediction model includes: merchant data samples of merchants, the merchant data samples including one or more types of merchant information samples;

[0173] The preset prediction model predicts the delivery order indicator based on the merchant data of the target merchant; the merchant data of the target merchant is obtained in the following manner:

[0174] If the target merchant does not have merchant information of the target type, merchant information of the target type of the target merchant is generated based on merchant information of the target type of similar merchants similar to the target merchant; wherein the similar merchants and the target merchant are located in the same business district and provide the same category of goods.

[0175] In some examples, the multiple candidate user AOIs matched by the target merchant are obtained in the following manner:

[0176] Taking the geographical location of the target merchant as the center, the geographical range within the preset distance threshold is divided into multiple planes;

[0177] The candidate user AOIs in each plane are acquired in parallel to obtain multiple candidate user AOIs matching the target merchant.

[0178] In some examples, the determination module 44 determines the delivery range of the target merchant based on the optimal user AOI combination, including:

[0179] The geographical boundaries of the outermost user AOIs in the optimal AOI combination are connected to obtain the delivery range of the target merchant.

[0180] The implementation process of the functions and effects of each module in the above-mentioned distribution range generation device is specifically described in the implementation process of the corresponding steps in the above-mentioned distribution range generation method, and will not be repeated here.

[0181] Accordingly, an embodiment of this specification also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the aforementioned embodiment of the method for generating a delivery range.

[0182] Accordingly, an embodiment of this specification also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the embodiment of the method for generating a delivery range are implemented.

[0183] Accordingly, an embodiment of this specification also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of an embodiment of a method for generating a delivery range when the computer program is executed by a processor.

[0184] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this specification. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0185] The above embodiments can be applied to one or more computer devices, where the computer device is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions. The hardware of the computer device includes but is not limited to a microprocessor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0186] The computer device may be any electronic product that can interact with a user, such as a personal computer, a tablet computer, a smart phone, a personal digital assistant (PDA), a game console, an interactive network television (IPTV), a smart wearable device, etc.

[0187] The computer device may also include a network device and / or a user device, wherein the network device includes, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of hosts or network servers.

[0188] The network where the computer device is located includes but is not limited to the Internet, wide area network, metropolitan area network, local area network, virtual private network (VPN), etc.

[0189] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0190] The steps of the various methods above are divided only for the purpose of clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent; adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this application.

[0191] Although this specification includes many specific implementation details, these should not be interpreted as limiting the scope of any invention or the scope of protection claimed, but are mainly used to describe the features of specific embodiments of specific inventions. Certain features described in multiple embodiments within this specification may also be implemented in combination in a single embodiment. On the other hand, the various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. In addition, although features may work in certain combinations as described above and even initially claimed as such, one or more features from the claimed combination may be removed from the combination in some cases, and the claimed combination may point to a sub-combination or a variation of the sub-combination.

[0192] The phrases "specific examples" or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of this specification. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0193] Other embodiments of the present invention will readily occur to those skilled in the art upon consideration of the present invention and practice of the invention claimed herein. This specification is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of this specification and include common knowledge or customary techniques in the art not claimed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present invention being indicated by the following claims.

[0194] It should be understood that the present description is not limited to the exact structure that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present description is limited only by the appended claims.

[0195] The above description is only a preferred embodiment of this specification and is not intended to limit this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of this specification should be included in the scope of protection of this specification.

Claims

1. A method for generating a delivery range, the method comprising: Obtain a target merchant for which a delivery range is to be generated and multiple candidate user AOIs matching the target merchant; Obtain the distance between each candidate user's AOI and the target merchant; With maximizing the total delivery order size of the target merchant as the optimization goal, the optimization model selects from each candidate user AOI to solve the optimal user AOI combination that meets the optimization goal and the constraints. The constraints include: if the target user AOI is not selected, other user AOIs that are farther away are not selected. Based on the optimal user AOI combination, the delivery range of the target merchant is determined.

2. The method according to claim 1, wherein the plurality of candidate user AOIs matching the target merchant are obtained by: Divide the preset geographical range corresponding to the target merchant into multiple planes, and obtain the candidate user AOI in each plane to obtain multiple candidate user AOIs matching the target merchant; The constraints include: In the case that the target user AOI is not selected, other user AOIs that are farther away in the plane to which the target user AOI belongs are not selected.

3. The method according to claim 1 or 2, wherein the delivery order indicator comprises: Delivery order quantity; The optimal optimization model includes a mixed integer programming model; The decision variables of the mixed integer programming model include binary variables for representing whether to select the candidate user AOI; The objective function of the mixed integer programming model includes: maximizing the total delivery order size of the target merchant, where the total delivery order size refers to the sum of the products of the binary variables of each candidate user AOI in multiple candidate user AOIs and the number of delivery orders; The constraints of the mixed integer programming model include: for multiple candidate user AOIs sorted from near to far based on their distance from the target merchant, if the value of the binary variable of the s-th user AOI indicates that the user AOI was not selected, then the binary variables of the s+1-th and all subsequent user AOIs with a farther distance are all set to the value indicating that the user AOI was not selected.

4. The method according to claim 1, further comprising: For each candidate user AOI, a preset prediction model is used to predict the delivery order index between the target merchant and the candidate user AOI; The constraint condition also includes: the predicted delivery order index is within a preset index range.

5. The method according to claim 4, wherein the delivery order indicators include: Number of delivery orders, delivery time, delivery distance and delivery efficiency; The prediction model includes a hybrid expert model, which is used to predict each of the delivery order indicators based on the input merchant data of the target merchant, the historical order data and basic attribute data of the candidate user AOI.

6. The method according to claim 4, wherein the training data of the preset prediction model comprises: Merchant data samples of merchants, wherein the merchant data samples include one or more types of merchant information samples; The preset prediction model predicts the delivery order indicator based on the merchant data of the target merchant; the merchant data of the target merchant is obtained in the following manner: If the target merchant does not have merchant information of the target type, merchant information of the target type of the target merchant is generated based on merchant information of the target type of similar merchants similar to the target merchant; wherein the similar merchants and the target merchant are located in the same business district and provide the same category of goods.

7. The method according to claim 1, wherein the plurality of candidate user AOIs matched by the target merchant are obtained by: Taking the geographical location of the target merchant as the center, the geographical range within the preset distance threshold is divided into multiple planes; The candidate user AOIs in each plane are acquired in parallel to obtain multiple candidate user AOIs matching the target merchant.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

9. A computer program product comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images