Article distribution method, device and equipment and computer readable medium

Abstract

The invention discloses an article distribution method, device and equipment and a computer readable medium, and relates to the technical field of logistics transportation. According to one specific embodiment, the method comprises the steps that an order area and an order area access sequence serve as training parameters, a delivery path model is obtained through training, and the delivery path model is used for outputting a delivery path; order area vectors are constructed through the order area data and the delivery data in the order area, and the order area vectors in the delivery area are aggregated to generate delivery area feature vectors; and taking the distribution area feature vectors and the distribution paths of the distribution areas as training data, performing training to obtain a connection path model, and outputting connection paths of a plurality of distribution areas by the connection path model so as to provide distribution articles for the distribution areas according to connection points in the connection paths. According to the embodiment, normal distribution of articles can be guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of logistics and transportation technology, and in particular to a method, device, equipment and computer-readable medium for distributing items. Background Art

[0002] Recently, the boom in e-commerce has driven the rapid development of the traditional logistics industry, bringing many development opportunities, but also some new challenges. In the logistics and distribution process, the last mile from the distribution center to the customer is usually the most expensive, so it is also considered the most challenging part.

[0003] Currently, the docking end transports the items from the distribution site to the docking point, and the distribution end only needs to distribute the items within the distribution area. In this way, the distribution end does not need to spend a lot of time traveling back and forth between the distribution area and the distribution site.

[0004] In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art: the fixed setting of the docking point requires the delivery end to pick up items at the docking point multiple times, which affects the normal delivery of the items. Summary of the invention

[0005] In view of this, embodiments of the present invention provide a method, apparatus, device and computer-readable medium for delivering items, which can ensure the normal delivery of items.

[0006] To achieve the above object, according to one aspect of an embodiment of the present invention, a method for distributing items is provided, comprising:

[0007] The order area and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output a delivery path;

[0008] Constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector;

[0009] The delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, and the docking path model outputs docking paths of multiple delivery areas to provide delivery items for the delivery areas according to the docking points in the docking paths.

[0010] Before taking the order area of ​​the delivery end and the order of visiting the order area as training parameters, the method further includes:

[0011] The delivery address of the order is converted into longitude and latitude coordinates, and the longitude and latitude coordinates are spatially clustered to determine the order area.

[0012] The delivery path model is obtained by training the order area of ​​the delivery end and the order of visiting the order area as training parameters, including:

[0013] The delivery path model is trained by taking the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, and the order of visiting the order area as training parameters.

[0014] The delivery path model is obtained by training the order area of ​​the delivery end and the order of visiting the order area as training parameters, including:

[0015] The delivery path model is trained by taking the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, the delivery influencing parameters and the order of visiting the order area as training parameters.

[0016] The step of constructing an order area vector using order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector, includes:

[0017] The order area data includes the order volume of the order area and the order volume of the connection point;

[0018] The delivery data in the order area includes the time point when the delivery end arrives at the order area, the movement time of the delivery end between the order areas, and the stay time of the delivery end in the order area.

[0019] The method further comprises:

[0020] Obtaining a docking point in a delivery area according to the docking path;

[0021] The order delivery time is determined based on the distance between the docking point and the delivery end, the order area and the current time.

[0022] The method of using the delivery area feature vector and the delivery path of the delivery area as training data to train a connection path model includes:

[0023] The delivery area feature vector and the delivery path of the delivery area are used as training data, and the connection path model is obtained by training using the REINFORCE algorithm.

[0024] According to a second aspect of an embodiment of the present invention, there is provided a device for distributing items, comprising:

[0025] A distribution module, used for training a distribution path model using the order area and the order of visiting the order area as training parameters, wherein the distribution path model is used for outputting the distribution path;

[0026] A vector module, for constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector;

[0027] The docking module is used to use the delivery area feature vector and the delivery path of the delivery area as training data to train a docking path model, and the docking path model outputs the docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths.

[0028] According to a third aspect of an embodiment of the present invention, there is provided an electronic device for delivering items, comprising:

[0029] one or more processors;

[0030] a storage device for storing one or more programs,

[0031] When the one or more programs are executed by the one or more processors, the one or more processors implement the above method.

[0032] According to a fourth aspect of an embodiment of the present invention, there is provided a computer-readable medium on which a computer program is stored, and when the program is executed by a processor, the method as described above is implemented.

[0033] One embodiment of the above invention has the following advantages or beneficial effects: the order area of ​​the delivery end and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output the delivery path; the order area vector is constructed through the order area data and the data in the order area, and the order area vector in the delivery area is aggregated to generate a delivery area feature vector; the delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, and the docking path model outputs the docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths. Since the docking path can meet the delivery paths of multiple delivery areas, the normal delivery of items in multiple delivery areas can be guaranteed.

[0034] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0036] Figure 1 is a schematic diagram of the main flow of a method for delivering items according to an embodiment of the present invention;

[0037] Figure 2is a schematic diagram of a process for determining order delivery time according to an embodiment of the present invention;

[0038] Figure 3 is a schematic diagram of the main structure of an apparatus for distributing items according to an embodiment of the present invention;

[0039] Figure 4 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0040] Figure 5 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0042] At present, in the last-mile delivery model, the service area of ​​the delivery station is generally divided into multiple delivery areas, and each delivery end is responsible for the delivery of items in one of the delivery areas. Specifically, after the delivery end obtains a batch of items at the delivery station, it transports them to the delivery area for delivery. After the delivery is completed, it returns to the delivery station to pick up the goods. However, when the distance between the delivery station and the delivery area is far, the delivery end frequently travels back and forth, which seriously affects the delivery efficiency.

[0043] The two-way delivery method is that the docking end is responsible for transporting the goods from the delivery site to the docking points in each delivery area, and the delivery end only needs to deliver the goods within the delivery area. In this way, the delivery end does not need to spend a lot of time traveling back and forth between the delivery area and the delivery site, thereby improving delivery efficiency and enhancing user experience.

[0044] The docking points are mostly set based on manual experience. Since the departure time of the docking end is often uncertain, and the delivery end cannot guarantee the exact time and driving route because the number of orders delivered each day is different. In this way, visiting fixed docking points in a fixed order may cause a lot of time and resource waste. Therefore, the fixed setting of docking points does not improve the punctuality of item delivery. Therefore, the fixed setting of docking points requires the delivery end to pick up items at the docking points multiple times, affecting the normal delivery of items.

[0045] In order to solve the problem that the delivery end picks up items at the docking point multiple times, affecting the normal delivery of items, the following technical solutions in the embodiments of the present invention can be adopted.

[0046] See also Figure 1 , Figure 1 FIG. 1 is a schematic diagram of the main flow of the method for delivering items according to an embodiment of the present invention, wherein a connection path is determined based on the delivery path. Figure 1 As shown in 100, the following steps are specifically included:

[0047] S101. Using the order area and the order of visiting the order area as training parameters, a delivery path model is obtained through training. The delivery path model is used to output a delivery path.

[0048] In the embodiment of the present invention, the order area is used as the basis for delivering the items. The order area is obtained based on the conversion of the delivery address of the order. That is, the delivery address of the order is converted into longitude and latitude coordinates, and the longitude and latitude coordinates are spatially clustered to determine the order area.

[0049] The order area can be a residential area, an office building, etc. The order area can be determined by analyzing the distribution characteristics of buildings in the physical world.

[0050] Usually, the delivery end will leave an order area only after completing the delivery of orders in that area. Therefore, mapping the delivery address of an order to the order area can improve the accuracy of the subsequent delivery route prediction of the delivery end based on the granularity of the order area.

[0051] In one embodiment of the present invention, the delivery end is used to deliver the items in the order to the delivery address. As an example, the delivery end includes a delivery vehicle. The delivery vehicle can carry the items in the order. As another example, the delivery end includes a delivery robot, and the delivery robot carries the items in the order.

[0052] In one embodiment of the present invention, the order area of ​​the delivery end and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output the delivery path. That is, the order area of ​​the delivery end is used as the input of the delivery path model, and the order of visiting the order area of ​​the delivery end is used as the output of the delivery path model to train the delivery path model.

[0053] See Table 1, which is an example of order data. The order destination address in Table 1 is the delivery address. The order area can be determined based on the delivery address.

[0054] Table 1

[0055]

[0056] See Table 2, which is an example of a delivery order at the delivery end. Based on the order number in Table 2, Table 1 can be queried to determine the order area at the delivery end.

[0057] Table 2

[0058] Order Number Delivery terminal identification Order Status time xx01563201xx 21xxxx deliver goods 11-12 09:30:30

[0059] See Table 3, Table 3 Delivery end trajectory data example. The order in which the delivery end visits the order area is determined based on the delivery end trajectory.

[0060] Table 3

[0061] Delivery terminal identification time longitude latitude 21xxxx 12-12 08:30:30 37.50 121.39

[0062] As an example, a delivery path model is trained based on the XGBoost model. That is, the order area of ​​the delivery end is used as the input of the XGBoost model, and the order in which the delivery end visits the order area is used as the output of the XGBoost model to train the delivery path model.

[0063] The XGBoost model is an excellent machine learning algorithm that is suitable for various prediction tasks, including regression and classification. The XGBoost model can be used to predict the delivery path of a delivery segment.

[0064] Specifically, the order area number is used as the input feature of the XGBoost model, and the delivery path is generated based on the historical orders. For example, the delivery path includes the order of the order area number, thereby training the XGBoost model.

[0065] During the model training process, the XGBoost model processes the input features. The XGBoost model uses the gradient boosting tree technology to iteratively train multiple weak learners to eventually obtain a powerful delivery path model.

[0066] At the same time, the delivery path model can be tested by using orders to evaluate the delivery path model. It can be observed that the delivery path model has obvious improvements in indicators such as edit distance and accuracy, proving its strong ability in delivery path planning.

[0067] Specifically, the delivery path model reduces the edit distance by 45% and improves the accuracy by 161% compared with the distance-based greedy algorithm.

[0068] In one embodiment of the present invention, the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, and the order in which the delivery end visits the order area are used as training parameters to train a delivery path model.

[0069] In order to improve the accuracy of predicting the delivery path, the training parameters can be based on the order area of ​​the delivery end and the order of the delivery end visiting the order area. The number of orders in the order area and the distance between the delivery end and the order area can also be used as training data. For the number of orders in the order area, in order to improve the delivery efficiency, the orders in the order area with a large number of orders can be given priority; for the distance between the delivery end and the order area, the order area with a short distance to the delivery end can be given priority, which helps to improve the delivery efficiency.

[0070] The number of orders in the order area and the distance between the delivery end and the order area are used as training data. The trained delivery path model can predict the delivery path based on the number of orders and distance in the order area, further improving the efficiency of delivering items in the order.

[0071] In one embodiment of the present invention, the timeliness of the items in the delivery order at the delivery end is considered to be affected by other factors. The delivery influencing parameters are used to characterize the factors affecting the delivery. As an example, the delivery influencing parameters include traffic influencing parameters and / or weather influencing factors.

[0072] The order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, the delivery impact parameters and the order in which the delivery end visits the order area are used as training parameters to train the delivery path model.

[0073] The delivery influencing parameters include traffic influencing parameters and / or weather influencing factors in historical orders. The delivery influencing parameters are used as training parameters, so that the delivery path model will also consider the traffic and / or weather at the prediction time during the delivery path prediction process, thereby improving the accuracy of the predicted delivery path.

[0074] In one embodiment of the present invention, the delivery path model specifically outputs the order in which the delivery end visits the order area. The order in which the delivery end visits the order area is combined with the movement time of the delivery end between the order areas and the stay time of the delivery end in the order area to obtain the delivery path of the delivery end. The delivery path includes the delivery time for the delivery end to reach the order area.

[0075] As an example, the length of time that the delivery end moves between order areas is determined by the average length of time that the delivery end moves between order areas in historical orders. The length of time that the delivery end stays in the order area is determined based on the type of order area and the number of orders in the order area. The types of order areas include residential areas, industrial areas, mixed commercial and residential areas, and shopping malls. For example, the model can be used to predict the length of time that the delivery end stays in the order area.

[0076] S102: construct an order area vector using the order area data and the delivery data in the order area, and aggregate the order area vectors in the delivery area to generate a delivery area feature vector.

[0077] There are multiple order areas in the distribution area, and each order area has corresponding order area data. For the distribution side, there is also corresponding distribution data in the order area, that is, the distribution data of the distribution side in the order area. The order area vector can be constructed through the order area data and the distribution data in the order area. The order area vector represents the characteristics of the order area.

[0078] In one embodiment of the present invention, the order area data includes the order volume of the order area and the order volume of the docking point. The order volume of the order area is the number of orders with the same delivery address in a preset time period. The order volume of the docking point is the number of orders in the order area obtained from the docking point in the preset time period.

[0079] The delivery data in the order area includes the time when the delivery end arrives at the order area, the movement time of the delivery end between the order areas, and the length of time the delivery end stays in the order area.

[0080] There are multiple order areas in the delivery area. The order area vectors in the delivery area are aggregated to generate the delivery area feature vector. The order area vector represents the order area, and the delivery area feature vector represents the delivery area. Based on the order area vector, the path is represented at the granularity of the order area, which fully reflects the importance of the order area in the delivery of items.

[0081] S103. Using the distribution area feature vector and the distribution path of the distribution area as training data, a docking path model is obtained through training. The docking path model outputs docking paths of multiple distribution areas to provide distribution items for the distribution areas according to the docking points in the docking paths.

[0082] In one embodiment of the present invention, the delivery area feature vector and the delivery path of the delivery area may be used as training data to obtain a docking path model through training. The docking path model outputs the docking paths of multiple delivery areas.

[0083] As an example, the connection path model is trained using the encoder-decoder model based on the attention mechanism. That is, the feature vector of the delivery area and the delivery path of the delivery area are used as training data, and the connection path model is trained using the encoder-decoder model based on the attention mechanism.

[0084] The input of the encoder-decoder model based on the attention mechanism is the representation sequence of each delivery area, and the output is the docking paths of these delivery areas. The attention mechanism can focus on important order areas, that is, order areas with a large number of orders, when generating docking paths. After the docking path is determined, delivery items can be provided to the delivery end in the delivery area according to the docking points in the docking path.

[0085] In one embodiment of the present invention, the distribution area feature vector and the distribution path of the distribution area are used as training data, and the REINFORCE algorithm is used to train and obtain a docking path model, so as to further improve the efficiency of item distribution.

[0086] REINFORCE is a policy gradient algorithm whose goal is to update policy parameters by gradient ascent to reduce the average delivery time of expected docking orders.

[0087] Experiments have shown that compared with the traditional heuristic docking path generation algorithm, the REINFORCE algorithm can reduce the average order delivery time by 30%, showing extremely high effectiveness.

[0088] Specifically, the REINFORCE algorithm trains and optimizes the model by iteratively sampling and evaluating various docking paths. In each iteration, a batch of docking paths are generated and the delivery time of each docking path is calculated. Then, based on the performance of the docking path, the gradient of each docking path is calculated, and the parameters of the docking path model are updated using the gradient ascent method.

[0089] By gradually optimizing the model in the above way, it can better select the connection path, thereby reducing the delivery time of the order. Compared with the traditional heuristic algorithm, the REINFORCE algorithm can significantly improve the efficiency of item delivery.

[0090] See also Figure 2 That is 200, Figure 2 1 is a flow chart of determining the order delivery time according to an embodiment of the present invention. Specifically, the following steps are included:

[0091] S201. Obtain the docking points in the delivery area according to the docking route.

[0092] The docking path includes multiple docking points, and the docking points correspond to the delivery areas. As an example, the docking points correspond to the delivery areas one by one, that is, there is one docking point in each delivery area. Based on the docking path, the docking points in the delivery area can be known.

[0093] S202: Determine the order delivery time based on the distance between the docking point and the delivery end, the order area and the current time.

[0094] According to the distance between the docking point and the delivery end, we can know how long it takes for the delivery end to obtain the items in the order from the docking point. According to the geographical location determined by the order area of ​​the order, we can determine the delivery time from the docking point to the order area, and determine the order delivery time based on the current time.

[0095] As an example, the time it takes for the delivery end to obtain the items in the order from the docking point is 60 minutes; the delivery time from the docking point to the order area is 120 minutes, the current time is 13:00, and the confirmed order delivery time is 16:00.

[0096] exist Figure 2 In the embodiment, the order delivery time can be accurately predicted based on the docking path.

[0097] In the above embodiment, the order area of ​​the delivery end and the order of visiting the order area are used as training parameters to train a delivery path model, which is used to output the delivery path; the order area vector is constructed through the order area data and the data in the order area, and the order area vector in the delivery area is aggregated to generate a delivery area feature vector; the delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, which outputs the docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths. Since the docking path can meet the delivery paths of multiple delivery areas, the normal delivery of items in multiple delivery areas can be guaranteed.

[0098] Moreover, by adopting the technical solution in the embodiment of the present invention, the docking point can be flexibly determined based on factors such as the number of orders, traffic conditions, and weather conditions, so as to improve the delivery efficiency of the docking end. Moreover, the delivery path can also cope with different delivery scenarios to reduce unnecessary repeated journeys, greatly improve the delivery efficiency, and thus enhance the user experience.

[0099] See also Figure 3 , Figure 3 is a schematic diagram of the main structure of an apparatus for distributing items according to an embodiment of the present invention. The apparatus for distributing items can implement a method for distributing items, such as Figure 3 As shown in 300, the device for distributing items specifically includes:

[0100] A distribution module 301 is used to train a distribution path model using the order area and the order of visiting the order area as training parameters, and the distribution path model is used to output a distribution path;

[0101] A vector module 302, for constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector;

[0102] The docking module 303 is used to use the delivery area feature vector and the delivery path of the delivery area as training data to train a docking path model, and the docking path model outputs docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths.

[0103] In one embodiment of the present invention, the delivery module 301 is further used to convert the delivery address of the order into longitude and latitude coordinates, and spatially cluster the longitude and latitude coordinates to determine the order area.

[0104] In one embodiment of the present invention, the distribution module 301 is specifically used to train a distribution path model using the order area of ​​the distribution end, the number of orders in the order area, the distance between the distribution end and the order area, and the order in which the distribution end visits the order area as training parameters.

[0105] In one embodiment of the present invention, the delivery module 301 is specifically used to train a delivery path model using the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, delivery influencing parameters and the order in which the delivery end visits the order area as training parameters.

[0106] In one embodiment of the present invention, the order area data includes the order volume of the order area and the order volume of the docking point;

[0107] The delivery data in the order area includes the time point when the delivery end arrives at the order area, the movement time of the delivery end between the order areas, and the stay time of the delivery end in the order area.

[0108] In one embodiment of the present invention, the docking module 303 is further used to obtain a docking point in the delivery area according to the docking path;

[0109] The order delivery time is determined based on the distance between the docking point and the delivery end, the order area and the current time.

[0110] In one embodiment of the present invention, the docking module 303 is specifically configured to use the delivery area feature vector and the delivery path of the delivery area as training data, and adopt the REINFORCE algorithm to train and obtain a docking path model.

[0111] Figure 4 An exemplary system architecture 400 is shown to which the method for delivering items or the apparatus for delivering items according to the embodiments of the present invention may be applied.

[0112] like Figure 4 As shown, system architecture 400 may include terminal devices 401, 402, 403, network 404 and server 405. Network 404 is used to provide a medium for communication links between terminal devices 401, 402, 403 and server 405. Network 404 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0113] Users can use terminal devices 401, 402, 403 to interact with server 405 through network 404 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 401, 402, 403, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0114] The terminal devices 401 , 402 , and 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0115] The server 405 may be a server that provides various services, such as a backend management server (only an example) that provides support for shopping websites browsed by users using the terminal devices 401, 402, and 403. The backend management server may analyze and process the received data such as product information query requests, and feed back the processing results (such as target push information, product information - only an example) to the terminal device.

[0116] It should be noted that the method for delivering items provided in the embodiment of the present invention is generally executed by the server 405 , and accordingly, the device for delivering items is generally disposed in the server 405 .

[0117] It should be understood that Figure 4 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0118] Reference below Figure 5 , which shows a schematic diagram of the structure of a computer system 500 of a terminal device suitable for implementing an embodiment of the present invention. Figure 5 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0119] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the system 500 are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0120] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that a computer program read therefrom is installed into the storage section 508 as needed.

[0121] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the central processing unit (CPU) 501, the above-mentioned functions defined in the system of the present invention are executed.

[0122] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. 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 may include, but are not limited to: 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 the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal 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 computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0123] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0124] The modules involved in the embodiments of the present invention may be implemented in software or hardware. The modules described may also be arranged in a processor. For example, they may be described as follows: a processor includes a distribution module, a vector module, and a docking module. The names of these modules do not, in some cases, constitute limitations on the modules themselves. For example, the distribution module may also be described as "used to train a distribution path model using the order area and the order of visiting the order area as training parameters, and the distribution path model is used to output a distribution path".

[0125] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiment; or may exist independently without being assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device includes:

[0126] The order area and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output a delivery path;

[0127] Constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector;

[0128] The delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, and the docking path model outputs docking paths of multiple delivery areas to provide delivery items for the delivery areas according to the docking points in the docking paths.

[0129] According to the technical solution of the embodiment of the present invention, the order area of ​​the delivery end and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output the delivery path; the order area vector is constructed through the order area data and the data in the order area, and the order area vector in the delivery area is aggregated to generate a delivery area feature vector; the delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, and the docking path model outputs the docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths. Since the docking path can meet the delivery paths of multiple delivery areas, the normal delivery of items in multiple delivery areas can be guaranteed.

[0130] The above specific implementation methods do not constitute limitations on the scope of protection of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention. It should be noted that in the technical solution disclosed in this disclosure, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

Claims

1. A method for delivering items, characterized in that: include: The order area and the order of visiting the order area are used as training parameters to train a delivery path model, and the delivery path model is used to output a delivery path; Constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector; The delivery area feature vector and the delivery path of the delivery area are used as training data to train a docking path model, and the docking path model outputs docking paths of multiple delivery areas to provide delivery items for the delivery areas according to the docking points in the docking paths.

2. The method for distributing items according to claim 1, characterized in that: Before taking the order area of ​​the delivery end and the order of visiting the order area as training parameters, the method further includes: The delivery address of the order is converted into longitude and latitude coordinates, and the longitude and latitude coordinates are spatially clustered to determine the order area.

3. The method for distributing items according to claim 1, characterized in that: The delivery path model is obtained by training the order area of ​​the delivery end and the order of visiting the order area as training parameters, including: The delivery path model is trained by taking the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, and the order of visiting the order area as training parameters.

4. The method for distributing items according to claim 1, characterized in that: The delivery path model is obtained by training the order area of ​​the delivery end and the order of visiting the order area as training parameters, including: The delivery path model is trained by taking the order area of ​​the delivery end, the number of orders in the order area, the distance between the delivery end and the order area, the delivery influencing parameters and the order of visiting the order area as training parameters.

5. The method for distributing items according to claim 1, characterized in that: The step of constructing an order area vector using order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector, includes: The order area data includes the order volume of the order area and the order volume of the connection point; The delivery data in the order area includes the time point when the delivery end arrives at the order area, the movement time of the delivery end between the order areas, and the stay time of the delivery end in the order area.

6. The method for distributing items according to claim 1, characterized in that: The method further comprises: Obtaining a docking point in a delivery area according to the docking path; The order delivery time is determined based on the distance between the docking point and the delivery end, the order area and the current time.

7. The method for distributing items according to claim 1, characterized in that: The method of using the delivery area feature vector and the delivery path of the delivery area as training data to train a connection path model includes: The delivery area feature vector and the delivery path of the delivery area are used as training data, and the connection path model is obtained by training using the REINFORCE algorithm.

8. A device for distributing items, characterized in that: include: A distribution module, used for training a distribution path model using the order area and the order of visiting the order area as training parameters, wherein the distribution path model is used for outputting the distribution path; A vector module, for constructing an order area vector through order area data and delivery data in the order area, and aggregating the order area vectors in the delivery area to generate a delivery area feature vector; The docking module is used to use the delivery area feature vector and the delivery path of the delivery area as training data to train a docking path model, and the docking path model outputs the docking paths of multiple delivery areas to provide delivery items for the delivery area according to the docking points in the docking paths.

9. An electronic device for delivering items, characterized in that: include: one or more processors; a storage device 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 according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images