Order collection amount determination method and device, electronic equipment and readable storage medium

Abstract

The invention provides an order collection amount determination method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of logistics. The method comprises the steps: obtaining the collection feature data of a target collection station; obtaining a first graph adjacency matrix corresponding to the target collection station, wherein the first graph adjacency matrix is used for describing a distribution relationship between the target collection station and a plurality of collection stations; and predicting the collection order amount of the target collection station according to the collection feature data of the target collection station and the first graph adjacency matrix. According to the embodiment of the invention, the prediction accuracy of the single collection amount of the target collection station can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of logistics technology, and in particular to a method and device for determining the quantity of collection orders, an electronic device, and a computer-readable storage medium. Background Art

[0002] At present, with the rapid development of the e-commerce industry, the number of online shoppers is increasing, which further promotes the rapid development of the logistics supply chain. In the entire logistics supply chain, including warehousing, picking, transportation, and distribution, each link plays a vital role. This process is usually after the user places an order, the warehouse will produce it. The production mainly includes picking, packing and other links, and then it will be sent from the warehouse to the sorting center. The sorting center is mainly responsible for diverting the goods, and then transporting them to various distribution stations. The distribution station is mainly responsible for sending goods to customers, so the distribution station is also called a terminal. This article mainly uses the distribution station to represent it. For the traditional logistics industry, the main responsibility of the distribution station is to receive and send express delivery, and it is responsible for collecting and delivering express delivery from customers. In the entire logistics distribution process, the number of collection orders at the distribution station directly affects the number of each node in the entire logistics distribution, thereby affecting the reasonable resource planning and transportation network planning of each node in the logistics. In summary, the reasonable estimation of the collection volume of the distribution station plays an indispensable role in logistics construction. How to improve the accuracy of the estimation of the collection volume of the distribution station is a very important issue in logistics. Summary of the invention

[0003] The purpose of the present disclosure is to provide a method, device, electronic device and computer-readable storage medium for determining the number of collection orders, which can accurately estimate the number of collection orders at a collection station.

[0004] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0005] The embodiment of the present disclosure provides a method for determining the number of collection orders, including: obtaining collection feature data of a target collection station; obtaining a first graph adjacency matrix corresponding to the target collection station, wherein the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations;

[0006] The collection order quantity of the target collection station is predicted based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0007] In some embodiments, obtaining the collection characteristic data of the target collection station includes: obtaining the collection characteristic data corresponding to multiple moments in a first time period of the target collection station; wherein the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and the multiple collection stations in the first time period; wherein, predicting the collection order volume of the target collection station based on the collection characteristic data of the target collection station and the first graph adjacency matrix includes: determining the collection order volume of the target collection station at the next moment in the first time period based on the collection characteristic data of the target collection station and the first graph adjacency matrix.

[0008] In some embodiments, the multiple collection stations include a first collection station and a second collection station; wherein, obtaining the first graph adjacency matrix corresponding to the target collection station includes: determining multiple nodes in the first graph adjacency matrix according to the target collection station and the multiple collection stations; if within the first time period, the delivery frequency between the target collection station and the first collection station is greater than a preset threshold, then in the first graph adjacency matrix, the node corresponding to the target collection station is connected to the node corresponding to the first collection station; if within the first time period, the delivery frequency between the target collection station and the second collection station is less than or equal to the preset threshold, then in the first graph adjacency matrix, the node corresponding to the target collection station is not connected to the node corresponding to the second collection station.

[0009] In some embodiments, the first time period includes a first moment, and the first moment is the last moment in the first time period; wherein, determining the number of collection orders of the target collection station at the next moment in the first time period based on the collection feature data of the target collection station and the first graph adjacency matrix includes: performing graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment to obtain a first graph convolution feature; predicting the number of collection orders of the target collection station at the next moment in the first time period based on the first graph convolution feature.

[0010] In some embodiments, the first time period also includes a second moment, and the second moment is the previous moment of the first moment; wherein, predicting the collection order quantity of the target collection station at the next moment of the first time period based on the first graph convolution feature includes: performing graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the second moment to obtain a second graph convolution feature; performing circular convolution processing on the first graph convolution feature and the second graph convolution feature to obtain a spatiotemporal convolution feature; predicting the collection order quantity of the target collection station at the next moment of the first time period based on the spatiotemporal convolution feature.

[0011] In some embodiments, before predicting the collection volume of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix, the method also includes: obtaining the collection feature data corresponding to multiple collection stations at multiple moments in a second time period; obtaining the second graph adjacency matrix corresponding to the multiple collection stations in the second time period, the second graph adjacency matrix being used to describe the distribution relationship between the multiple collection stations in the second time period; the target neural network determines the predicted collection volume of each collection station at the next moment in the second time period based on the collection feature data of the multiple collection stations and the second graph adjacency matrix; obtains the actual collection volume of the multiple collection stations at the next moment in the second time period; and trains the target neural network based on the predicted collection volume and actual collection volume of the multiple collection stations at the next moment in the second time period, so as to predict the collection volume of the target collection station based on the target neural network.

[0012] In some embodiments, the first time period includes a third moment; wherein, obtaining the collection characteristic data of the target collection station at the third moment includes: determining the time characteristics of the third moment according to the date, month, week number, whether it is a weekend, whether it is a big promotion, or whether it is a holiday of the third moment; determining the spatial characteristics of the target collection station according to the province, city level, distribution area, number of employees, and number of road areas of the target collection station; determining the collection volume characteristics of the target collection station at the first N moments before the third moment, where N is an integer greater than or equal to 1; determining the collection characteristic data of the target collection station at the third moment according to the time characteristics of the third moment, the spatial characteristics of the target collection station, and the collection volume characteristics of the target collection station at the first N moments before the third moment.

[0013] An embodiment of the present disclosure provides a device for determining the number of collection orders, including: a collection feature data acquisition module, a graph adjacency matrix determination module and a collection order quantity determination module.

[0014] Among them, the collection feature data acquisition module is used to obtain the collection feature data of the target collection station; the graph adjacency matrix determination module can be used to obtain the first graph adjacency matrix corresponding to the target collection station, and the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations; the collection order quantity determination module can be used to predict the collection order quantity of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0015] An embodiment of the present disclosure proposes an electronic device, which includes: a memory and a processor; the memory is used to store program instructions; the processor calls the program instructions stored in the memory to implement any of the above-mentioned methods for determining the number of collection orders.

[0016] An embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the method for determining the number of collection orders as described in any one of the above items is implemented.

[0017] The embodiment of the present disclosure provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above-mentioned method for determining the number of collection orders.

[0018] The method, device, electronic device and computer-readable storage medium for determining the number of collection orders provided by the embodiments of the present disclosure combine the collection feature data of the target collection station and the relationship network data between distribution stations when predicting the number of collection orders of a target collection station. Through the distribution relationship data between the target collection station and other collection stations, the number of collection orders of the target collection station can be accurately predicted, so as to effectively guide resource planning and reduce logistics operating costs.

[0019] It is to be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0021] Figure 1 A schematic diagram showing an exemplary system architecture that can be applied to the method or device for determining the number of collected orders in an embodiment of the present disclosure.

[0022] Figure 2 The present invention is a flowchart of a method for determining the number of collection orders according to an exemplary embodiment.

[0023] Figure 3 The figure is a flow chart of a method for determining collection feature data according to an exemplary embodiment.

[0024] Figure 4 The present invention is a flowchart of a method for determining a graph adjacency matrix according to an exemplary embodiment.

[0025] Figure 5 It is a schematic diagram of a graph for describing a delivery relationship according to an exemplary embodiment.

[0026] Figure 6 It is a schematic diagram of a graph adjacency matrix according to an exemplary embodiment.

[0027] Figure 7 The present invention is a flowchart of a method for determining the number of collection orders according to an exemplary embodiment.

[0028] Figure 8 The present invention is a flowchart of a method for determining the number of collection orders according to an exemplary embodiment.

[0029] Fig. 9 It is a schematic diagram of a neural network for predicting the number of collected orders according to an exemplary embodiment.

[0030] Fig.10 A neural network training method for determining the number of collected orders is shown according to an exemplary embodiment.

[0031] Fig.11 It is a block diagram of a device for determining the number of collection orders according to an exemplary embodiment.

[0032] Fig.12 A schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0033] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be comprehensive and complete and will fully convey the concepts of the example embodiments to those skilled in the art. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted.

[0034] The features, structures or characteristics described in the present disclosure may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0035] The accompanying drawings are only schematic diagrams of the present disclosure, and the same reference numerals in the drawings represent the same or similar parts, so their repeated description will be omitted. Some block diagrams shown in the accompanying drawings do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0036] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and steps, nor must they be executed in the order described. For example, some steps can be decomposed, and some steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0037] In the description of the present application, unless otherwise specified, " / " means "or", for example, A / B can mean A or B. "And / or" in this article is only a kind of association relationship describing the associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, "at least one" means one or more, and "plurality" means two or more. The words "first", "second", etc. do not limit the quantity and execution order, and the words "first", "second", etc. do not limit them to be different; the terms "comprising", "including" and "having" are used to express the meaning of open inclusion and mean that in addition to the listed elements / components / etc., there may be other elements / components / etc.

[0038] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0039] In the relevant technology, the estimation technology of the order quantity of the collection station includes qualitative method and quantitative method. The qualitative method is to make judgments based on the experience of management personnel and operation personnel. This method is still common in most logistics, and is usually summarized and calculated in the form of offline documents. The quantitative method is to build models through statistical methods and machine learning methods. Statistical methods include linear regression, nonlinear regression, etc. Machine learning methods include tree regression, neural network, etc. This method obtains necessary information based on historical data. Compared with the simple and crude qualitative method, it considers more comprehensive influencing factors, the rationality of modeling, and the quality of data directly affect the accuracy and reliability of order quantity estimation.

[0040] In the related art, the estimation results of the order quantity estimation method of the collection station are often greatly deviated. Therefore, the technical solution provided in this embodiment can provide a method for accurately estimating the order quantity of the collection station, and the specific details can be referred to the embodiments of this application.

[0041] The exemplary embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0042] Figure 1 A schematic diagram showing an exemplary system architecture that can be applied to the method or device for determining the number of collected orders in an embodiment of the present disclosure.

[0043] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0044] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, wearable devices, virtual reality devices, smart homes, etc.

[0045] The server 105 may be a server that provides various services, such as a background management server that provides support for devices operated by users using the terminal devices 101, 102, and 103. The background management server may analyze and process the received request data, and feed back the processing results to the terminal device.

[0046] The server can be an independent physical server or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as basic cloud computing services such as big data and artificial intelligence platforms, etc. The present disclosure does not impose any restrictions on this.

[0047] The server 105 may, for example, obtain the collection feature data of the target collection station; the server 105 may, for example, obtain the first graph adjacency matrix corresponding to the target collection station, the first graph adjacency matrix being used to describe the delivery relationship between the target collection station and multiple collection stations; the server 105 may, for example, predict the collection order volume of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0048] It should be understood that Figure 1 The number of terminal devices, networks and servers in the figure is only for illustration. The server 105 may be a physical server or may be composed of multiple servers. According to actual needs, it may have any number of terminal devices, networks and servers.

[0049] Under the above system architecture, a method for determining the number of collection orders is provided in an embodiment of the present disclosure. The method can be executed by any electronic device with computing and processing capabilities.

[0050] Figure 2 The method provided in the embodiment of the present disclosure can be executed by any electronic device with computing and processing capabilities. For example, the method can be executed by the above-mentioned Figure 1 The server or terminal device in the embodiment may be executed, or the server and the terminal device may be executed together. In the following embodiments, the server is used as the execution subject for example, but the present disclosure is not limited to this.

[0051] Reference Figure 2 The method for determining the number of collected orders provided in the embodiment of the present disclosure may include the following steps.

[0052] Step S202, obtaining the collection characteristic data of the target collection station.

[0053] The collection station may refer to the station where waybills are collected and delivered during the logistics process.

[0054] The collection feature data may refer to data that can describe the collection characteristics of the collection station, such as the time characteristics, spatial characteristics and collection volume characteristics of the collection station. Among them, the time characteristics of the collection station may refer to the time characteristics corresponding to the collection station at a certain moment, such as the specific date, month, week number, whether it is a weekend, whether it is a big promotion, whether it is a holiday, etc., which may affect the collection volume. Spatial characteristics may refer to characteristics such as the province where the collection station is located, the city level, the distribution area, the number of employees and the number of road areas. The collection volume characteristics may refer to the collection volume characteristics of the N moments before a certain moment.

[0055] Step S204, obtaining a first graph adjacency matrix corresponding to the target collection station, where the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations.

[0056] In some embodiments, a first graph adjacency matrix ( Figure 7 As shown, Figure 7 Is a description Figure 5 The graph adjacency matrix of the graph shown).

[0057] Step S206, predicting the number of collection orders of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0058] The technical solution provided in the above embodiment combines the collection feature data of the target collection station and the relationship network data between the distribution stations when predicting the collection order volume of the target collection station. Through the distribution relationship data between the target collection station and other collection stations, the collection order volume of the target collection station can be accurately predicted, so as to effectively guide resource planning and reduce logistics operating costs.

[0059] Figure 3 The figure is a flow chart of a method for determining collection feature data according to an exemplary embodiment.

[0060] refer to Figure 3 The above-mentioned method for determining collection characteristic data may include the following steps.

[0061] Step S302, obtaining the collection feature data corresponding to multiple time points of the target collection station within the first time period.

[0062] In some embodiments, the first time period may include multiple moments, and the moment may refer to a day, a week, etc.

[0063] In some embodiments, obtaining the collection feature data corresponding to multiple time points of the target collection station within the first time period may refer to the collection feature data corresponding to each day (or each month) of the target collection station in a month.

[0064] In some embodiments, the first time period may include a third time.

[0065] Then the following will take the third moment as an example to explain how to obtain the collection feature data of the target collection station at the third moment.

[0066] In some embodiments, the time characteristics of the third moment can be determined based on the date, month, week number, whether it is a weekend, whether it is a big sale, or whether it is a holiday; the spatial characteristics of the target collection station can be determined based on the province, city level, distribution area, number of employees, and number of road areas of the target collection station; the collection volume characteristics of the target collection station at the first N moments before the third moment can be determined, where N is an integer greater than or equal to 1; and then the collection feature data of the target collection station at the third moment can be determined based on the time characteristics of the third moment, the spatial characteristics of the target collection station, and the collection volume characteristics of the first N moments before the third moment.

[0067] The above process may be understood specifically by referring to the following embodiments.

[0068] In some embodiments, basic data of the target delivery station may be obtained, and the basic data may include historical waybills of the delivery station, basic information of the delivery station, etc.

[0069] Among them, the distribution station historical waybills, including the basic information of the waybills, can be shown in the following table:

[0070]

[0071] The basic information of the distribution station may include the basic information of the station, which may be shown in the following table:

[0072] Fields illustrate Delivery station ID Unique number of the delivery station Province name The name of the province where the distribution station is located City Name The name of the city where the delivery station is located City level The level of the city where the distribution station is located Coverage area Distribution station coverage area Number of employees Number of employees at distribution stations Number of road areas Number of distribution station road areas

[0073] In some embodiments, the above basic data may be cleaned, and the data cleaning mainly includes basic operations such as field type conversion, null value processing, and field deduplication.

[0074] In some embodiments, after data cleaning is completed on the basic data, feature extraction may be performed on the basic data.

[0075] Feature extraction, as shown in the following table, includes date features, mainly month, date, day of the week, weekend, promotion, holiday, etc.; distribution station information, mainly province, city level, distribution area, number of employees, number of road areas. Recent collection statistics, including recent sales, mean, standard deviation, proportion of zero orders, etc.; among the above features, discrete features need to be one-hot encoded, as shown in the following table.

[0076]

[0077] The above characteristics are the characteristics of the distribution station every day, recorded here as Where i represents the i-th distribution station, j represents the j-th feature of the distribution station, and t represents the date sequence number, such as t=1, which represents the first time point.

[0078] Step S304, obtaining a first graph adjacency matrix corresponding to the target collection station, where the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations within the first time period;

[0079] Step S306, determining the number of collection orders of the target collection station at the next moment in the first time period according to the collection feature data of the target collection station and the first graph adjacency matrix.

[0080] Figure 4 The present invention is a flowchart of a method for determining a graph adjacency matrix according to an exemplary embodiment.

[0081] In some embodiments, the plurality of pick-up stations may include a first pick-up station and a second pick-up station.

[0082] This embodiment will take the first collection station and the second collection station as examples to explain the method for generating a graph adjacency matrix.

[0083] refer to Figure 4 , the above-mentioned graph adjacency matrix determination method may include the following steps.

[0084] Step S402: determining a plurality of nodes in the first graph adjacency matrix according to the target collection station and the plurality of collection stations.

[0085] like Figure 5 As shown, multiple nodes in the first graph adjacency matrix can be determined based on the target collection station and multiple collection stations that have a delivery relationship with the target collection station. Figure 5 As shown, the target collection station can be node A in the adjacency matrix of the first graph, and other collection stations can be node B, node C, node D, node E, etc. in the adjacency matrix of the first graph.

[0086] Step S404: If the delivery frequency between the target collection station and the first collection station is greater than a preset threshold value within the first time period, the node corresponding to the target collection station is connected to the node corresponding to the first collection station in the first graph adjacency matrix.

[0087] like Figure 5 As shown, if the configuration frequency of the collection station corresponding to node A and the collection station corresponding to node B in the first time period is greater than a preset threshold, node A and node B are connected.

[0088] Step S406: If the delivery frequency between the target collection station and the second collection station is less than or equal to the preset threshold value within the first time period, the node corresponding to the target collection station is not connected to the node corresponding to the second collection station in the first graph adjacency matrix.

[0089] like Figure 5As shown, if the delivery frequency between the collection station corresponding to node A and the collection station corresponding to node C is less than or equal to the preset threshold, node A and node C will not be connected.

[0090] Specifically, if Figure 5 As shown in the figure, A, B, C, D, and E represent the distribution stations, and the lines between the nodes indicate that there are orders to be delivered between the distribution stations. From the above figure, we can get the adjacency matrix of the graph, as shown in Figure 6 shown.

[0091] Generate adjacency matrix logic, count the average daily delivery volume in the first time period (such as the last 3 months), and determine whether two delivery stations are connected. Then connect the two nodes.

[0092] The technical solution provided by the above embodiment can generate a graph adjacency matrix that can describe the spatial distribution relationship between the target collection station and multiple collection stations based on the distribution relationship between the target collection station and multiple collection stations, so as to predict the single collection volume of the target collection station based on the graph adjacency matrix.

[0093] Figure 6 The present invention is a flowchart of a method for determining the number of collection orders according to an exemplary embodiment.

[0094] In some embodiments, the first time period may include a first moment, and the first moment is the last moment in the first time period.

[0095] refer to Figure 6 The above-mentioned method for determining the number of collection orders may include the following steps.

[0096] Step S602: Perform graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment to obtain a first graph convolution feature.

[0097] In some embodiments, the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment can be convolved through a graph convolutional network (such as GCN) to obtain a first graph convolution feature.

[0098] Step S604: predict the number of collection orders of the target collection station at the next moment in the first time period according to the first graph convolution feature.

[0099] The technical solution provided by the above embodiment can convolute and fuse the first graph adjacency matrix describing the spatial distribution relationship in the target collection with the feature data that can describe the collection characteristics of the target collection station at the first moment, so as to determine a first convolution feature that can describe both the spatial collection characteristics of the target collection station and the collection characteristics of the target collection station itself. Since the first graph convolution feature can describe both the collection data characteristics of the target collection station itself and the spatial collection characteristics of the target collection station, predicting the single collection volume of the target collection station based on the first convolution feature can improve the prediction accuracy.

[0100] Figure 8 The present invention is a flowchart of a method for determining the number of collection orders according to an exemplary embodiment.

[0101] In some embodiments, the first time period may include a first moment and a second moment, the second moment may be the previous moment of the first moment (such as the second to last day in the first time period), and the first moment may be the last moment in the first time period (such as the last day in the first time period).

[0102] refer to Figure 8 The above-mentioned method for determining the number of collection orders may include the following steps.

[0103] Step S802: Perform graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment to obtain a first graph convolution feature.

[0104] like Fig. 9 As shown, the first graph adjacency matrix A and the collection feature data corresponding to the target collection station at the first moment (such as Fig. 9 The Xt in (Xt) is processed by graph convolution through the graph convolution network GCN to obtain the first graph convolution feature.

[0105] In some embodiments, the first graph adjacency matrix A and the collection feature data corresponding to the target collection station at the first moment can be processed by graph convolution through a multi-layer graph convolution network to obtain the first graph convolution feature, which is not limited in this application.

[0106] Step S804 performs graph convolution processing on the first graph adjacency matrix and the collected feature data corresponding to the second moment to obtain a second graph convolution feature.

[0107] like Fig. 9 As shown, the first graph adjacency matrix A and the collection feature data corresponding to the target collection station at the second moment (such as Fig. 9 Xt-1) in is processed by graph convolution through the graph convolution network GCN to obtain the second graph convolution feature.

[0108] In some embodiments, the first graph adjacency matrix A and the collection feature data corresponding to the target collection station at the second moment can be processed by graph convolution through a multi-layer graph convolution network to obtain the second graph convolution feature, which is not limited in this application.

[0109] In some embodiments, the first time period may also include at least one other moment (such as the fifth moment and the sixth moment), and the other moments may also refer to the processing methods of the first moment and the second moment to obtain graph convolution features. This application does not impose any restrictions on this.

[0110] Step S806: Perform circular convolution processing on the first graph convolution feature and the second graph convolution feature to obtain spatiotemporal convolution features.

[0111] like Fig. 9 As shown, the first convolution feature and the second convolution feature can be subjected to cyclic convolution processing through a recurrent neural network (such as GRU (Gated Recurrent Unit) or LSTM (Long Short-Term Memory)) to obtain spatiotemporal convolution features.

[0112] Step S808, predicting the number of collection orders of the target collection station at the next moment in the first time period based on the spatiotemporal convolution features.

[0113] In some embodiments, the number of collection orders at the target collection station at the next moment in the first time period can be predicted based on the above-mentioned spatiotemporal convolution features.

[0114] The following will Fig. 9 The network shown is described in detail.

[0115] Fig. 9 The input of the model shown includes the target delivery station feature matrix X and the adjacency matrix A. The model can be expressed as The overall model structure is as follows Fig. 9 shown.

[0116] The core of the model mainly consists of two parts, graph convolution and RNN layer.

[0117] (1) Graph Convolutional Network (GCN)

[0118] Considering the large number of distribution stations, a single-layer GCN is used in this application (however, a multi-layer GCN can also be used. The technology in this field can be set according to actual needs, and this application does not limit this). Its expression is Among them, X is the characteristic matrix of the distribution station, Among them I N is a diagonal matrix, is the degree matrix of the adjacency matrix A, which is defined as W is the parameter in the network that needs to be trained.

[0119] (2) RNN layer

[0120] In this layer, GRU is selected in the present invention, and the output of the above graph convolution GCN will be used as the input of GRU.

[0121] A representation of the relationship between the inputs and outputs of the entire model It can be expressed as follows:

[0122] u t =σ(W u [f(A,X t ),h t-1 ]+b u ) (1)

[0123] r t =σ(W r [f(A,X t ),h t-1 ]+b r ) (2)

[0124] c t =tanh(W c [f(A,X t ),(r t ,h t-1 )]+b c ) (3)

[0125] h t =u t *h t-1 +(1-u t )*c t (4)

[0126] Among them, the number of collection orders can be predicted based on the above ht

[0127] In practical applications, you can use calculate Fig. 9 The average accuracy of the model shown can be determined based on actual application results. Fig. 9 The model shown is about 5% better than the traditional machine learning model that does not introduce spatial features.

[0128] Fig.10 A neural network training method for determining the number of collected orders is shown according to an exemplary embodiment.

[0129] refer to Fig.10 The training method of the neural network for determining the number of collected orders may include the following steps.

[0130] Step S1002, obtaining collection feature data corresponding to multiple collection stations at multiple times in the second time period.

[0131] In some embodiments, the first time period and the second time period may be the same or different, and this application does not limit this.

[0132] In some embodiments, the first time period may be after the second time period.

[0133] In some embodiments, the plurality of pick-up stations may include the target pick-up station.

[0134] Step S1004, obtaining a second graph adjacency matrix corresponding to the plurality of collection stations in the second time period, where the second graph adjacency matrix is ​​used to describe the delivery relationship between the plurality of collection stations in the second time period.

[0135] Step S1006: The target neural network determines the predicted number of collection orders for each collection station at the next moment in the second time period based on the collection feature data of the multiple collection stations and the second graph adjacency matrix.

[0136] The target neural network can be constructed by a graph neural network and a recurrent neural network that can perform graph convolution processing on the graph adjacency matrix and the collected feature data (such as Fig. 9 shown).

[0137] In some embodiments, the target neural network determines the predicted collection order quantity of each collection station at the next moment in the second time period according to the collection feature data of multiple collection stations and the second graph adjacency matrix. Figure 6 to Figure 9 The process shown is not described in detail in this application.

[0138] Step S1008, obtaining the actual number of collection orders at the multiple collection stations at the next moment in the second time period.

[0139] Step S1010, training a target neural network based on the predicted collection order quantities and actual collection order quantities of multiple collection stations at the next moment in the second time period, so as to predict the collection order quantity of the target collection station based on the target neural network.

[0140] The above embodiment shows a training method for training a target neural network, through which the prediction accuracy of the target neural network for predicting the single collection volume of the target collection station can be improved.

[0141] Based on the same inventive concept, the embodiments of the present disclosure also provide a device for determining the number of collected orders, such as the following embodiments. Since the principle of solving the problem in the device embodiment is similar to that in the above method embodiment, the implementation of the device embodiment can refer to the implementation of the above method embodiment, and the repeated parts will not be repeated.

[0142] Fig.11 is a block diagram of a device for determining the number of collection orders according to an exemplary embodiment. Fig.11 The device 1100 for determining the number of collected orders provided in the embodiment of the present disclosure may include: a collection feature data acquisition module 1101, a graph adjacency matrix determination module 1102 and a collection order quantity determination module 110.

[0143] Among them, the collection feature data acquisition module 1101 can be used to obtain the collection feature data of the target collection station; the graph adjacency matrix determination module 1102 can be used to obtain the first graph adjacency matrix corresponding to the target collection station, and the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations; the collection order quantity determination module 1103 can be used to predict the collection order quantity of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0144] It should be noted that the above-mentioned collection feature data acquisition module 1101, graph adjacency matrix determination module 1102 and collection order quantity determination module 1103 correspond to S202 to S206 in the method embodiment, and the above-mentioned modules and corresponding steps implement the same examples and application scenarios, but are not limited to the contents disclosed in the above-mentioned method embodiment. It should be noted that the above-mentioned modules as part of the device can be executed in a computer system such as a set of computer executable instructions.

[0145] In some embodiments, the collection feature data acquisition module 1101 may include: a feature data acquisition unit at multiple moments; the collection order quantity determination module 1103 may include: a collection order quantity prediction unit.

[0146] Among them, the characteristic data acquisition unit for multiple moments is used to obtain the collection characteristic data corresponding to the target collection station at multiple moments within the first time period; the collection order quantity prediction unit can be used to determine the collection order quantity of the target collection station at the next moment in the first time period based on the collection characteristic data of the target collection station and the first graph adjacency matrix; wherein, the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations within the first time period.

[0147] In some embodiments, the plurality of collection stations include a first collection station and a second collection station; wherein the graph adjacency matrix determination module 1102 may include: a node determination unit, a connection judgment unit, and a disconnection judgment unit.

[0148] Among them, the node determination unit can be used to determine multiple nodes in the first graph adjacency matrix according to the target collection station and multiple collection stations; the connection judgment unit can be used to connect the node corresponding to the target collection station with the node corresponding to the first collection station in the first graph adjacency matrix if the delivery frequency between the target collection station and the first collection station is greater than a preset threshold within the first time period; the non-connection judgment unit can be used to not connect the node corresponding to the target collection station with the node corresponding to the second collection station in the first graph adjacency matrix if the delivery frequency between the target collection station and the second collection station is less than or equal to the preset threshold within the first time period.

[0149] In some embodiments, the first time period includes a first moment, and the first moment is the last moment in the first time period; wherein, the collected order quantity prediction unit may include: a first graph convolution processing subunit and a first collected order quantity determination subunit.

[0150] Among them, the first graph convolution processing subunit can be used to perform graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment to obtain the first graph convolution feature; the first collection order quantity determination subunit can be used to predict the collection order quantity of the target collection station at the next moment in the first time period based on the first graph convolution feature.

[0151] In some embodiments, the first time period also includes a second moment, which is the previous moment of the first moment; wherein the first determination subunit for the number of collected orders may include: a second graph convolution processing grandson unit, a circular convolution processing grandson unit and a second determination grandson unit for the number of collected orders.

[0152] Among them, the second graph convolution processing grandson unit can be used to perform graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the second moment to obtain the second graph convolution feature; the circular convolution processing grandson unit can be used to perform circular convolution processing on the first graph convolution feature and the second graph convolution feature to obtain the spatiotemporal convolution feature; the second grandson unit for determining the collection order quantity can be used to predict the collection order quantity of the target collection station at the next moment in the first time period based on the spatiotemporal convolution feature.

[0153] In some embodiments, the device for determining the number of collected orders may include: a second collection feature data acquisition module, a second graph neighbor acquisition module, a neural network prediction module, an actual collection quantity acquisition module and a training module.

[0154] Among them, the second collection feature data acquisition module can be used to obtain the collection feature data corresponding to multiple collection stations at multiple times within the second time period before predicting the collection order volume of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix; the second graph neighbor acquisition module can be used to obtain the second graph adjacency matrix corresponding to multiple collection stations in the second time period, and the second graph adjacency matrix is ​​used to describe the distribution relationship between multiple collection stations in the second time period; the neural network prediction module can be used for the target neural network to determine the predicted collection order volume of each collection station at the next moment in the second time period based on the collection feature data of multiple collection stations and the second graph adjacency matrix; the actual collection volume acquisition module can be used to obtain the actual collection order volume of multiple collection stations at the next moment in the second time period; the training module can be used to train the target neural network based on the predicted collection order volume and the actual collection order volume of multiple collection stations at the next moment in the second time period, so as to predict the collection order volume of the target collection station according to the target neural network.

[0155] In some embodiments, the first time period includes a third moment; wherein the collection feature data acquisition module 1101 includes: a time feature determination unit, a space feature determination unit, a collection quantity feature determination unit and a collection feature data determination unit.

[0156] Among them, the time feature determination unit can be used to determine the time features of the third moment according to the date, month, week number, whether it is a weekend, whether it is a big promotion, or whether it is a holiday at the third moment; the spatial feature determination unit can be used to determine the spatial features of the target collection station according to the province, city level, distribution area, number of employees and number of road areas of the target collection station; the collection volume feature determination unit can be used to determine the collection volume features of the target collection station at the previous N moments before the third moment, where N is an integer greater than or equal to 1; the collection feature data determination unit can be used to determine the collection feature data of the target collection station at the third moment according to the time features of the third moment, the spatial features of the target collection station and the collection volume features of the target collection station at the previous N moments before the third moment.

[0157] Since the functions of the device 1100 have been described in detail in the corresponding method embodiments, the present disclosure will not elaborate on them here.

[0158] The modules and / or units and / or sub-units and / or grandchildren units described in the embodiments of the present application may be implemented by software or hardware. The modules and / or units and / or sub-units and / or grandchildren units described may also be arranged in a processor. The names of these modules and / or units and / or sub-units and / or grandchildren units do not, in certain circumstances, constitute limitations on the modules and / or units and / or sub-units and / or grandchildren units themselves.

[0159] 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 application. 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.

[0160] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0161] Fig.12 The structure diagram of the electronic device suitable for implementing the embodiment of the present disclosure is shown. It should be noted that: Fig.12 The electronic device 1200 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0162] like Fig.12 As shown, the electronic device 1200 includes a central processing unit (CPU) 1201, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1202 or a program loaded from a storage part 1208 into a random access memory (RAM) 1203. In the RAM 1203, various programs and data required for the operation of the electronic device 1200 are also stored. The CPU 1201, the ROM 1202, and the RAM 1203 are connected to each other via a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0163] The following components are connected to the I / O interface 1205: an input section 1206 including a keyboard, a mouse, etc.; an output section 1207 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1208 including a hard disk, etc.; and a communication section 1209 including a network interface card such as a LAN card, a modem, etc. The communication section 1209 performs communication processing via a network such as the Internet. A drive 1210 is also connected to the I / O interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1210 as needed, so that a computer program read therefrom is installed into the storage section 1208 as needed.

[0164] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part 1209, and / or installed from a removable medium 1211. When the computer program is executed by a central processing unit (CPU) 1201, the above-mentioned functions defined in the system of the present application are executed.

[0165] It should be noted that the computer-readable storage medium shown in the present disclosure 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 application, 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 application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. 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. The computer-readable signal medium may also be any computer-readable storage medium other than a computer-readable storage medium that can 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 storage 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.

[0166] As another aspect, the present application also provides a computer-readable storage medium, which may be included in the device described in the above embodiment; or it may exist independently without being assembled into the device. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed by a device, the device can implement functions including: obtaining the collection feature data of the target collection station; obtaining the first graph adjacency matrix corresponding to the target collection station, the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations; predicting the collection order quantity of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

[0167] According to one aspect of the present application, a computer program product or a computer program is provided, the computer program product or the computer program including computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method provided in various optional implementations of the above-mentioned embodiments.

[0168] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software, or by combining software with necessary hardware. Therefore, the technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computing device (which can be a personal computer, a server, a mobile terminal, or a smart device, etc.) to execute the method according to the embodiment of the present disclosure, for example Figure 2 to Figure 4 , Figure 7 , Figure 8 ,or Fig.10 One or more of the steps shown.

[0169] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the disclosure disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not applied for by the present disclosure. The specification and examples are to be regarded as exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

[0170] It should be understood that the present disclosure is not limited to the detailed structures, drawings or implementations shown herein, but rather the present disclosure is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method for determining the number of collected orders, characterized in that: include: Obtaining collection characteristic data of the target collection station; Obtaining a first graph adjacency matrix corresponding to the target collection station, where the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations; The collection order volume of the target collection station is predicted based on the collection feature data of the target collection station and the first graph adjacency matrix.

2. The method according to claim 1, characterized in that: Obtain the target collection station's collection characteristic data, including: Acquire collection feature data corresponding to multiple time points of the target collection station within the first time period; The first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and the multiple collection stations within the first time period; The method of predicting the number of collection orders of the target collection station according to the collection feature data of the target collection station and the first graph adjacency matrix includes: The number of collection orders of the target collection station at the next moment in the first time period is determined according to the collection feature data of the target collection station and the first graph adjacency matrix.

3. The method according to claim 2, characterized in that: The plurality of collection stations include a first collection station and a second collection station; wherein obtaining a first graph adjacency matrix corresponding to the target collection station includes: Determine a plurality of nodes in the first graph adjacency matrix according to the target collection station and the plurality of collection stations; If the delivery frequency between the target collection station and the first collection station is greater than a preset threshold value during the first time period, connecting the node corresponding to the target collection station with the node corresponding to the first collection station in the first graph adjacency matrix; If, within the first time period, the delivery frequency between the target collection station and the second collection station is less than or equal to the preset threshold, the node corresponding to the target collection station is not connected to the node corresponding to the second collection station in the first graph adjacency matrix.

4. The method according to claim 2, characterized in that: The first time period includes a first moment, and the first moment is the last moment in the first time period; wherein determining the number of collection orders of the target collection station at the next moment in the first time period according to the collection feature data of the target collection station and the first graph adjacency matrix includes: Performing graph convolution processing on the first graph adjacency matrix and the collection feature data corresponding to the target collection station at the first moment to obtain a first graph convolution feature; Predict the number of collection orders of the target collection station at the next moment in the first time period according to the first graph convolutional features.

5. The method according to claim 4, characterized in that: The first time period also includes a second moment, and the second moment is a previous moment of the first moment; wherein predicting the number of collection orders of the target collection station at the next moment of the first time period according to the first graph convolution feature includes: Performing graph convolution processing on the first graph adjacency matrix and the collected feature data corresponding to the second moment to obtain a second graph convolution feature; Performing circular convolution processing on the first graph convolution feature and the second graph convolution feature to obtain spatiotemporal convolution features; The number of collection orders of the target collection station at the next moment in the first time period is predicted according to the spatiotemporal convolution features.

6. The method according to claim 1, characterized in that: Before predicting the collection order quantity of the target collection station according to the collection feature data of the target collection station and the first graph adjacency matrix, the method further includes: Acquire collection characteristic data corresponding to multiple collection stations at multiple times in the second time period; Obtaining a second graph adjacency matrix corresponding to the plurality of collection stations in the second time period, where the second graph adjacency matrix is ​​used to describe the distribution relationship between the plurality of collection stations in the second time period; The target neural network determines the predicted collection order quantity of each collection station at the next moment in the second time period according to the collection feature data of the plurality of collection stations and the second graph adjacency matrix; Obtaining the actual number of collection orders of the multiple collection stations at the next moment in the second time period; The target neural network is trained based on the predicted collection order quantities and actual collection order quantities of the multiple collection stations at the next moment in the second time period, so as to predict the collection order quantity of the target collection station based on the target neural network.

7. The method according to claim 2, characterized in that: The first time period includes a third moment; wherein obtaining the collection characteristic data of the target collection station at the third moment includes: Determine the time feature of the third moment according to the date, month, week number, whether it is a weekend, whether it is a big sale, or whether it is a holiday of the third moment; Determine the spatial characteristics of the target collection station according to the province, city level, distribution area, number of employees and number of road areas of the target collection station; Determine the collection volume characteristics of the target collection station at the previous N moments before the third moment, where N is an integer greater than or equal to 1; The collection characteristic data of the target collection station at the third moment is determined based on the time characteristics of the third moment, the spatial characteristics of the target collection station, and the collection volume characteristics of the target collection station at the first N moments before the third moment.

8. A device for determining the number of collected orders, characterized in that: include: A collection feature data acquisition module is used to obtain the collection feature data of the target collection station; A graph adjacency matrix determination module, used to obtain a first graph adjacency matrix corresponding to the target collection station, wherein the first graph adjacency matrix is ​​used to describe the distribution relationship between the target collection station and multiple collection stations; A collection order quantity determination module is used to predict the collection order quantity of the target collection station based on the collection feature data of the target collection station and the first graph adjacency matrix.

9. An electronic device, characterized in that: include: Memory; as well as A processor coupled to the memory, the processor being used to execute the method for determining the number of collection orders as described in any one of claims 1-7 based on instructions stored in the memory.

10. A computer-readable storage medium having program instructions stored thereon, which, when executed by a processor, implements the method for determining the number of collection orders as described in any one of claims 1 to 7.

Images