CN114239977A

CN114239977A - Method, device, equipment and storage medium for determining estimated delivery time length

Info

Publication number: CN114239977A
Application number: CN202111574557.2A
Authority: CN
Inventors: 胡启万; 张智标; 茹强
Original assignee: Beijing Sankuai Online Technology Co Ltd
Current assignee: Beijing Sankuai Online Technology Co Ltd
Priority date: 2021-12-21
Filing date: 2021-12-21
Publication date: 2022-03-25

Abstract

The application discloses a method, a device, equipment and a storage medium for determining pre-estimated delivery time length, and belongs to the technical field of internet. The method comprises the following steps: acquiring time length estimation reference information, and performing feature extraction on the time length estimation reference information based on a feature extraction model to obtain feature information corresponding to the time length estimation reference information; respectively inputting the characteristic information into time length estimation models corresponding to a plurality of distribution stages to obtain the stage time length corresponding to each distribution stage; and determining the estimated delivery time length based on the corresponding stage time length of each delivery stage. Compared with the method for determining the delivery time length only according to the delivery distance, the method for determining the estimated delivery time length comprehensively considers the time consumption corresponding to each stage in the whole delivery process, and considers various factors possibly influencing the delivery time length, so that the estimated delivery time length can be determined more accurately.

Description

Method, device, equipment and storage medium for determining estimated delivery time length

Technical Field

The present application relates to the field of internet technologies, and in particular, to a method, an apparatus, a device, and a storage medium for determining a pre-estimated delivery duration.

Background

With the development of network information, the influence of logistics distribution business on the life of people is more and more serious. The logistics distribution business can be a takeout business, an express business, a flash business and the like. When a user selects corresponding delivery service in the application programs of the logistics delivery service platform, the application program of the platform displays estimated delivery time required by the completion of delivery to the user, and simultaneously, a deliverer is arranged to complete delivery.

Currently, the estimated delivery duration is mainly calculated according to the delivery distance between a delivery place and a receiving place.

In fact, the process from the order placement by the user to the delivery of the goods or the merchandise to the receiving place by the delivery person can be divided into a plurality of links. For example, after a user purchases a meal item of a merchant a in an application of a takeout platform, the user can be divided into four links, i.e., a delivery task is received by a rider, the rider arrives at a store of the merchant a, the rider gets the meal item, and the rider goes to a destination from the store. Therefore, the method for determining the estimated delivery time length through the delivery distance only determines the time length required by one link in the whole delivery process, and does not consider the time lengths required by other links in the whole delivery process, so that the determined estimated delivery time length is inaccurate.

Disclosure of Invention

The embodiment of the application provides a method, a device, equipment and a storage medium for determining estimated delivery time length, and can solve the problem that the estimated delivery time length is inaccurate in the prior art. The technical scheme is as follows:

in a first aspect, a method for determining a pre-estimated delivery duration is provided, the method comprising:

acquiring duration estimation reference information, wherein the duration estimation reference information comprises delivery order information, environment information corresponding to the delivery order information and distributor distribution information;

based on the characteristic extraction model, carrying out characteristic extraction on the time length estimation reference information to obtain characteristic information corresponding to the time length estimation reference information;

respectively inputting the characteristic information into time length estimation models corresponding to a plurality of distribution stages to obtain the stage time length corresponding to each distribution stage;

and determining the estimated delivery time length based on the corresponding stage time length of each delivery stage.

In one possible implementation, the plurality of delivery stages includes a stage from a user placing an order to a delivery person receiving an order, a stage from a delivery person receiving an order to a delivery person arriving at a delivery site, a stage from a delivery person arriving at a delivery site to a delivery person getting goods, and a stage from a delivery person getting goods to a delivery person arriving at a receiving site.

In one possible implementation, the delivery order information includes at least one of a consignor name, a goods quantity, a goods unit price, an actual payment amount of the user, a delivery cost, a location information of the delivery place, and a location information of the receiving place;

the environment information comprises at least one of weather information and traffic information;

the distributor distribution information includes at least one of the number of distributors whose distance from the delivery site does not exceed the distance threshold, and location information of distributors whose distance from the delivery site does not exceed the distance threshold.

In one possible implementation, the feature extraction model includes a first feature extraction module and a second feature extraction module, and the first feature extraction module and the second feature extraction module are machine learning networks of different algorithms;

based on the feature extraction model, feature extraction is carried out on the time length estimation reference information to obtain feature information corresponding to the time length estimation reference information, and the feature information comprises the following steps:

based on a first feature extraction module, performing feature extraction on the time length estimation reference information to obtain first sub-feature information corresponding to the time length estimation reference information;

based on a second feature extraction module, performing feature extraction on the time length estimation reference information to obtain second sub-feature information corresponding to the time length estimation reference information;

and determining characteristic information corresponding to the duration estimation reference information based on the first sub-characteristic information and the second sub-characteristic information.

In a possible implementation manner, the feature extraction is performed on the duration pre-estimation reference information based on a first feature extraction module to obtain first sub-feature information corresponding to the duration pre-estimation reference information, and the method includes:

determining a first parameter vector based on a numerical parameter and a non-numerical parameter in the duration pre-estimation reference information;

inputting the first parameter vector into a first feature extraction module to obtain first sub-feature information corresponding to the duration pre-estimation reference information;

based on the second feature extraction module, the feature extraction is performed on the time length pre-estimation reference information to obtain second sub-feature information corresponding to the time length pre-estimation reference information, and the method comprises the following steps:

carrying out split-point discretization on the numerical parameters in the duration estimation reference information to obtain discretized numerical parameters;

determining a second parameter vector based on the non-numerical parameter in the discretization numerical parameter and the time length pre-estimation reference information;

and inputting the second parameter vector into a second feature extraction module to obtain second sub-feature information corresponding to the duration pre-estimation reference information.

In a possible implementation manner, determining an estimated delivery duration based on a phase duration corresponding to each delivery phase includes:

determining the weight corresponding to each distribution stage based on the duration estimation reference information and the weight calculation model;

and determining the estimated delivery time length based on the corresponding stage time length and the weight of each delivery stage.

In a possible implementation manner, determining a weight corresponding to each delivery phase based on the duration estimation reference information and the weight calculation model includes:

and inputting the second parameter vector into the weight calculation model to obtain the weight corresponding to each distribution stage.

determining a correction time-supplementing duration based on the duration pre-estimation reference information and the correction time-supplementing calculation model;

and determining the estimated delivery time length based on the corrected time-compensating time length and the stage time length corresponding to each delivery stage.

In a possible implementation manner, determining the modified time-complementing duration based on the duration pre-estimation reference information and the modified time-complementing calculation model includes:

and inputting the second parameter vector into a correction time-supplementing calculation model, and determining the correction time-supplementing duration.

In a second aspect, an apparatus for determining an estimated delivery duration is provided, the apparatus comprising:

the system comprises an acquisition module, a storage module and a display module, wherein the acquisition module is used for acquiring duration estimation reference information, and the duration estimation reference information comprises delivery order information, environment information corresponding to the delivery order information and distributor distribution information;

the characteristic extraction module is used for extracting characteristics of the time length estimation reference information based on the characteristic extraction model to obtain characteristic information corresponding to the time length estimation reference information;

the stage duration estimation module is used for respectively inputting the characteristic information into duration estimation models corresponding to the plurality of distribution stages to obtain stage duration corresponding to each distribution stage;

and the determining module is used for determining the estimated delivery time length based on the stage time length corresponding to each delivery stage.

the feature extraction module is configured to:

In one possible implementation manner, the feature extraction module is configured to:

the feature extraction module is configured to:

In one possible implementation manner, the determining module is configured to:

In a third aspect, a computer device is provided that includes a processor and a memory having stored therein at least one instruction that is loaded and executed by the processor to perform operations performed by a method of determining a pre-estimated delivery duration.

In a fourth aspect, a computer-readable storage medium is provided that has at least one instruction stored therein, the instruction being loaded and executed by a processor to perform operations performed by a method of determining a pre-estimated delivery duration.

In a fifth aspect, a computer program product is provided, the computer program product comprising computer program code which, when executed by a computer device, causes the computer device to perform the method of the first aspect and possible implementations thereof.

The technical scheme provided by the embodiment of the application has the following beneficial effects:

according to the scheme provided by the embodiment of the application, the distribution order information, the environment information corresponding to the distribution order information and the distributor distribution information can be obtained, and the characteristics of the information are extracted. According to the extracted characteristic information, the stage duration required by each distribution stage in the whole distribution process can be determined, and further the estimated distribution time can be determined. Compared with the method for determining the delivery time length only according to the delivery distance, the method for determining the estimated delivery time length comprehensively considers the time consumption corresponding to each stage in the whole delivery process, and considers various factors possibly influencing the delivery time length, so that the estimated delivery time length can be determined more accurately.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a flowchart of a method for determining an estimated delivery duration according to an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of delivery order information according to an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a distributor distribution information provided by an embodiment of the present application;

fig. 4 is a flowchart of a method for determining an estimated delivery duration according to an embodiment of the present application;

fig. 5 is a flowchart of a method for determining an estimated delivery duration according to an embodiment of the present application;

fig. 6 is a flowchart of a method for determining an estimated delivery duration according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of an apparatus for determining an estimated delivery duration according to an embodiment of the present application;

fig. 8 is a block diagram of a computer device according to an embodiment of the present disclosure.

Detailed Description

To make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

The method for determining the estimated delivery time provided by the embodiment of the application can be applied to the management of the logistics distribution platform on the logistics distribution business, and specifically can be used for determining the estimated delivery time of the logistics distribution business. The logistics platform may be a platform that provides logistics services, such as a take-away platform, a courier platform, a flash platform, and so on. When a user selects a corresponding delivery service in an application program of a logistics delivery platform, the user can see the estimated delivery duration corresponding to the delivery service, for example, after the user selects a commodity to be purchased in the application program of a certain takeout platform, the takeout platform displays the estimated delivery duration required for completing the delivery service to the user through the application program.

Next, the method for determining the estimated delivery duration provided by the present application is described in detail by taking a delivery service corresponding to a takeaway platform as an example in the embodiments of the present application, and other logistics delivery services are similar to the foregoing, and are not described in detail in the embodiments of the present application.

The embodiment of the application provides a method for determining estimated delivery duration, and an execution subject of the method can be a server. The server may include a processor, memory, and communication components, among others.

The processor may be a Central Processing Unit (CPU), a System on Chip (SoC), or the like. The processor may be configured to obtain time length estimation reference information, perform feature extraction on the time length estimation reference information, determine a stage time length corresponding to each delivery stage, determine a weight corresponding to each delivery stage, determine a modified time compensation time length, determine an estimated delivery time length, and the like.

The Memory may be various volatile memories or nonvolatile memories, such as a SSD (Solid State Disk), a DRAM (Dynamic Random Access Memory), and the like. The memory may be used to store pre-stored data, intermediate data and result data in determining the estimated delivery duration. For example, the time length estimation reference information, the characteristic information corresponding to the estimation reference information, the stage time length corresponding to each distribution stage, the weight corresponding to each distribution stage, the corrected time compensation time length, the estimated distribution time length, and the like.

The communication means may be a wired network connector, a WiFi (Wireless Fidelity) module, a bluetooth module, a cellular network communication module, etc. The communication component may be used for data transmission with other devices, and the other devices may be other servers, operation terminals, and the like. For example, the communication component may be configured to send the phase duration corresponding to each delivery phase to the terminal corresponding to the user, may be configured to send the estimated delivery duration to the terminal corresponding to the user, may be configured to receive duration estimation reference information sent by the terminal corresponding to the user, and the like.

The user can select the commodity to be purchased from the application program of the take-away platform, and after the selection is completed, the user can click the 'settlement' option to enter an order information confirmation interface. In the order information confirmation interface, the user may select or add corresponding shipping information, such as the location of the shipping location, the name of the recipient, the contact of the recipient, the expected delivery time of the recipient, and the like. After the delivery information is added, the application program can send a request for determining the estimated delivery duration to the server through the terminal of the user. After receiving the request and determining the estimated delivery time length corresponding to the order, the server can send the estimated delivery time length to the terminal of the user and display the estimated delivery time length to the user. The user may determine whether to make payment based on the estimated delivery duration. If the user does not accept the estimated delivery time, the user can quit the order information confirmation interface and end the order. If the user accepts the estimated delivery duration, the user can click the 'order submitting' option or the 'payment' option to enter a payment interface, and after the payment is finished, the user waits for the delivery of the commodities. After the user finishes payment, the takeaway platform can generate corresponding delivery order information again, and the server can determine the estimated delivery time again and display the estimated delivery time to the user through the application program. Next, the scheme is described by taking an example that the server determines the estimated delivery time after the user completes payment, and other situations are similar to the above, and are not described herein again.

The embodiment of the application provides a processing flow of the method for determining the estimated delivery duration as shown in fig. 1, which includes the following processing steps:

s101, acquiring duration estimation reference information.

The server can obtain the estimated time length reference information of the time length firstly when determining the estimated delivery time length. The duration estimation reference information may include delivery order information, environment information corresponding to the delivery order information, and distributor distribution information.

After the user places an order, the terminal used by the user can send the delivery order information to the server. The delivery order information may include at least one of a merchant name (shipper name), a goods name, a quantity of goods, a unit price of the goods, an amount actually paid by the user, a delivery fee, location information of the shipping place, and location information of the receiving place, as shown in fig. 2.

The environmental information may include at least one of weather information and traffic information.

The weather information may include weather of the administrative area where the shipping location is located, and weather of the administrative area where the receiving location is located. The server may store a string relation table corresponding to various weather in advance, as shown in table 1. After determining the position of the delivery location and the position of the receiving location from the delivery order information, the server determines administrative areas corresponding to the delivery location and the receiving location according to preset administrative area division rules, so that weather corresponding to the delivery location and the receiving location can be acquired.

Alternatively, as for the weather information, when the delivery location and the receipt location are in the same administrative area, the weather information may be the weather of the delivery location or the weather of the receipt location, and when a plurality of administrative areas are spanned between the delivery location and the receipt location, the weather information may be the weather of all administrative areas through which the delivery task is completed.

The traffic information may be road congestion between the origin and the destination. After the server acquires the position information of the delivery place and the position information of the receiving place, a plurality of delivery routes can be planned for the deliverers. The server may divide each delivery route into a plurality of unit road segments, which may be ten meters, twenty meters, fifty meters, and so on. The server may calculate the total congestion value of the distribution route according to the congestion values of a plurality of unit road segments belonging to the same distribution route, and the calculation method may be calculating an average value, calculating a median, calculating a mode, and the like. Finally, the server may determine the traffic information corresponding to the delivery order information according to the total congestion value of each delivery route, and the determining method may be calculating an average value, calculating a median, calculating a mode, and the like.

The distributor distribution information may include at least one of the number of distributors whose distance from the delivery site does not exceed the distance threshold, and location information of distributors whose distance from the delivery site does not exceed the distance threshold, as shown in fig. 3. After the server acquires the location information of the delivery place, the server may determine the number of current distributors in a circular area range with the delivery place as a center and a preset distance threshold as a radius, and determine the current location information of each distributor. Wherein the distance threshold may be 0.5 km, 1 km, 2 km, etc.

S102, extracting the characteristics of the time length estimation reference information based on the characteristic extraction model to obtain the characteristic information corresponding to the time length estimation reference information.

The feature extraction model may include a first feature extraction module and a second feature extraction module, and the first feature extraction module and the second feature extraction module may be machine learning Networks of different algorithms, for example, the first feature extraction module includes DNN (Deep Neural Networks ), BN (Batch Normalization standards) Networks and leak functions, the second feature extraction module is DCN (Deep & Cross Networks, Deep and Cross Networks), and so on.

After the server obtains the pre-estimated reference information, the numerical parameters and the non-numerical parameters in the pre-estimated reference information can be determined. The numerical parameters may include the quantity of the goods, the unit price of the goods, the amount actually paid by the user, the delivery cost, traffic information, and the number of deliverers who are not more than a distance threshold from the delivery location. Non-numeric parameters may include shipper name, item name, location information for the ship's place, weather information, location information for the shipper whose distance from the ship's place does not exceed a distance threshold, and so forth. The server can determine a first parameter vector and a second parameter vector according to the numerical parameter and the non-numerical parameter. The first and second parameter vectors are determined as follows:

first, determine the first parameter vector

The server can input all the non-numerical parameters into the feature mapping module respectively to determine the vector corresponding to each non-numerical parameter. The feature mapping module may be an Embedding module, which may use a low-dimensional vector to represent a discrete non-numeric parameter, for example, the server inputs the name of the shipper "some fast food restaurant" into the Embedding module, and may obtain a 10-dimensional vector [1,1,1,1,1,0,0,0 ]. The server may splice vectors corresponding to each non-numerical parameter to obtain a first sub-parameter vector, and a dimension of the first sub-parameter vector may be equal to a sum of dimensions of vectors corresponding to each non-numerical parameter.

After different non-numerical parameters are processed by the feature mapping module, the dimensions of the obtained vectors may be the same, for example, the server inputs the name of the shipper, the category of the goods, and the name of the goods into the Embedding module respectively, and generates a 10-dimensional vector after processing, or the server inputs the name of the shipper, the category of the goods, and the name of the goods into the Embedding module respectively, and generates a 32-dimensional vector after processing, and so on. Or, after different non-numerical parameters are processed by the feature mapping module, the dimensions of the obtained vectors may be different, for example, the server inputs the name of the shipper, the goods category, and the goods name into the Embedding module respectively, and the processed name of the shipper corresponds to a 10-dimensional vector, the processed name of the shipper corresponds to a 32-dimensional vector, and the processed name of the goods corresponds to a 64-dimensional vector, and so on. For the determination of the dimension of the vector output by the feature mapping module, the adjustment may be performed according to the size of the actual model, which is not described herein again.

The server may determine the second sub-parameter vector according to the numerical parameter. Each dimension in the second sub-parameter vector represents a numerical parameter, for example, the second sub-parameter vector is [2,28.8,3,7,15], which indicates that the number of goods is 2, the actual payment amount of the user is 28.8 yuan, the delivery cost is 3 yuan, the traffic information corresponds to a congestion value of 7 (i.e., congestion), and the number of current distributors is 15.

The server may splice the first sub-parameter vector and the second sub-parameter vector to obtain the first parameter vector.

Determining a second parameter vector

The server can carry out the discrete processing of quantile to numerical parameter to obtain the discrete numerical parameter, and the discrete processing of quantile of each numerical parameter can be independent each other. The process of the split-site discretization treatment can be as follows: for any numerical parameter, the server determines a target value range to which the numerical parameter belongs in a plurality of pre-stored value range ranges corresponding to the numerical parameter, determines a target character string corresponding to the target value range based on the corresponding relation between each pre-stored value range and the character string, and takes the target character string as the discretization numerical parameter corresponding to the numerical parameter.

The multiple value range corresponding to any numerical parameter can be determined according to the boundary point and the quantile point corresponding to the numerical parameter, wherein the boundary point and the quantile point are prestored in the server. For example, the server prestores boundary points corresponding to the delivery cost as 0 and 35, and corresponding quantiles as 5,10 and 20, and then determines that the value range corresponding to the delivery cost is [0,5), [5,10), [10,20 and [20,35], and further for example, the server prestores boundary points corresponding to the quantity of goods as 0 and 12, and corresponding quantiles as 3,6 and 9, and then determines that the value range corresponding to the quantity of goods as [0,3), [3,6), [6,9 ] and [9,12], and so on.

In the server, a character string corresponding to each value range is stored in advance. For example, the distribution fees correspond to ranges of values [0, 5], [5,10 ], [10,20], [20,35], the strings corresponding to ranges of values are PSFY _5, PSFY _10, PSFY _20, PSFY _35, respectively, and further, for example, the cargo quantities correspond to ranges of values [0,3 ], [3,6 ], [6,9 ], [9,12], the strings corresponding to ranges of values are HWSL _3, HWSL _6, HWSL _9, HWSL _12, respectively, and so on.

The classification point may be determined empirically by a developer, or may be determined after a certain number of samples are processed by a statistical method. The method for setting the quantiles can be an equidistant bucket method, the width between any two adjacent quantiles is fixed, namely, the value range is fixed, for example, the quantile discretization module corresponding to the distribution cost comprises four quantiles of 3,6, 9 and 12. Alternatively, the method for setting the quantiles may be an equal frequency bucket method, in this case, the quantiles may be determined by the server according to statistics on a plurality of numerical parameters belonging to the same type, for example, statistics on delivery costs in 30 delivery orders, where 10 belong to the interval [0, 5], 10 belong to the interval [5,10 ], and 10 belong to the interval [10,20], and thus, the quantiles corresponding to the delivery costs are determined as 5,10, and 20. Or, the method for setting the quantiles may be a model bucket method, and the server may pre-establish a quantile model and input a plurality of numerical parameters belonging to the same type into the quantile model, thereby determining a suitable quantile, for example, the server inputs the delivery cost of 50 delivery orders into the quantile model, and the quantile model determines 1, 3, 5, and 10 as the quantile corresponding to the delivery cost. The setting methods of the multi-component sites preset in the server may be the same or different, and are not described herein again.

The server can input each discretization numerical parameter into the feature mapping module respectively, and determine a vector corresponding to each discretization numerical parameter. The feature mapping module may be an Embedding module, which is not described herein again. The server can splice the vectors corresponding to each discretization numerical parameter to obtain a third sub-parameter vector.

The server may splice the first sub-parameter vector and the third sub-parameter vector, so as to obtain a second parameter vector.

The server may input the first parameter vector into the first feature extraction module, and obtain first sub-feature information (which may be the first feature vector) corresponding to the duration prediction reference information. The server may input the second parameter vector to the second feature extraction module, so as to obtain second sub-feature information (which may be a second feature vector) corresponding to the duration prediction reference information. The server may splice the first feature vector and the second feature vector to obtain a total feature vector, and the total feature vector may be regarded as feature information corresponding to the duration prediction reference information.

Alternatively, in step S102, in addition to the above-described processing manner of stitching the plurality of vectors, the plurality of vectors may be subjected to a bit alignment addition process, a bit alignment subtraction process, a number-of-lines-first multiplication and then stitching process, or the like.

S103, respectively inputting the characteristic information into the duration estimation models corresponding to the multiple distribution stages to obtain the stage duration corresponding to each distribution stage.

The process from the ordering of the user in the application program of the takeout platform to the delivery of the goods to the receiving place by the deliverer can be divided into a plurality of delivery stages. The multiple delivery stages may include a stage from a user placing an order to a deliverer receiving an order, a stage from a deliverer receiving an order to a deliverer reaching a target merchant (a delivery site), a stage from a deliverer reaching a target merchant (a delivery site) to a deliverer getting goods, a stage from a deliverer getting goods to a deliverer reaching a receiving site.

Aiming at the plurality of distribution stages, a time length estimation model corresponding to each distribution stage is trained in the server in advance. The server may input the feature information (i.e., the total feature vector) corresponding to the time length estimation reference information into the time length estimation model corresponding to each distribution stage, respectively, to obtain the stage time length corresponding to each distribution stage.

Alternatively, for the distribution service, in addition to the above-described processing manner of dividing into four distribution stages, the distribution service may be divided into two distribution stages, three distribution stages, six distribution stages, and the like. For example, in the case of a distribution service where a merchant makes its own distribution, a distributor arrives at the merchant before a user places an order, and in this case, the plurality of distribution stages include a stage from the user placing an order to the distributor to pick up goods (i.e., a stage from the user placing an order to the merchant to complete preparation of goods), and a stage from the distributor to the receiving place. For another example, for a cross-city delivery service, the plurality of delivery stages include a stage of placing an order from a user to a delivery location delivery person, a stage of getting a good from the delivery location delivery person to the delivery location delivery person, a stage of getting a good from the delivery location delivery person to a delivery location city transfer station, a stage of getting a good from the delivery location city transfer station to a receiving location city transfer station, a stage of getting a good from the receiving location city transfer station to the receiving location delivery person, and a stage of getting a good from the receiving location delivery person to the receiving location delivery person.

And S104, determining the estimated distribution time length based on the corresponding stage time length of each distribution stage.

After the server determines the stage duration corresponding to each distribution stage, the stage durations can be added to obtain the estimated distribution duration corresponding to the distribution service. For example, the server determines that the period of time from the order placement of the user to the order pickup of the distributor is 5 minutes, the period of time from the order pickup of the distributor to the arrival of the distributor at the target merchant is 6 minutes, the period of time from the arrival of the distributor at the target merchant to the pickup of the goods by the distributor is 2 minutes, and the period of time from the pickup of the goods by the distributor to the arrival of the distributor at the receiving place is 15 minutes, and then the estimated delivery period of time for completing the delivery service is 28(5+6+2+15) minutes.

By adopting the method for determining the pre-estimated delivery time length, the stage time length required by each delivery stage in the whole delivery process can be determined, the time consumption corresponding to each delivery stage in the whole delivery process is comprehensively considered, and the pre-estimated delivery time length corresponding to the completion of delivery service can be accurately determined.

The embodiment of the application provides a processing flow of the method for determining the estimated delivery duration as shown in fig. 4, and the method can calculate the weight corresponding to each delivery stage, and includes the following processing steps:

s201, acquiring duration estimation reference information.

The duration estimation reference information comprises delivery order information, environment information corresponding to the delivery order information and distributor distribution information. In this step, the duration prediction reference information obtained by the server is the same as the duration prediction reference information in step S101, and is not described herein again.

S202, extracting the characteristics of the time length estimation reference information based on the characteristic extraction model to obtain the characteristic information corresponding to the time length estimation reference information.

The feature extraction model may include a first feature extraction module and a second feature extraction module, and the first feature extraction module and the second feature extraction module may be machine learning networks of different algorithms. In this step, the characteristic information corresponding to the duration estimation reference information determined by the server is the same as the characteristic information in step S102, and is not described herein again.

S203, inputting the characteristic information into the duration estimation models corresponding to the multiple distribution stages respectively to obtain the stage duration corresponding to each distribution stage.

The distribution service can be divided into a plurality of distribution stages, and the server stores a duration estimation model corresponding to each distribution stage in advance. In this step, the phase duration corresponding to each delivery phase determined by the server is the same as the phase duration in step S103, and details are not repeated here.

S204, determining the weight corresponding to each distribution stage based on the duration estimation reference information and the weight calculation model.

The determination of the corresponding weight for each delivery phase may be as follows:

first, the server may determine the numerical parameters and the non-numerical parameters in the duration estimation reference information. Secondly, the server can perform discretization processing on the numerical parameter to obtain a discretization numerical parameter, and a second parameter vector is determined according to the discretization numerical parameter, the non-numerical parameter and the feature mapping model. Finally, the server inputs the second parameter vector into the weight calculation model, so that the weight corresponding to each distribution stage can be determined. The second parameter vector is determined in step S102, and is not described herein again.

S205, determining the estimated delivery time length based on the corresponding stage time length and the weight of each delivery stage.

After the server determines the stage duration and the weight corresponding to each distribution stage, the server can perform weighted summation on the stage duration to determine the estimated distribution duration corresponding to the distribution service. For example, the server determines that the period of time from the customer placing an order to the distributor receiving an order is 5 minutes, the weight value is 0.5, the period of time from the distributor receiving an order to the distributor reaching the target merchant is 6 minutes, the weight value is 0.8, the period of time from the distributor reaching the target merchant to the distributor getting the goods is 2 minutes, the weight value is 0.3, the period of time from the distributor getting the goods to the distributor reaching the receiving place is 15 minutes, and the weight value is 1.2, and then the estimated distribution period of time for completing the distribution service is 21.58 (5: 0.5+ 6: 0.8+ 2: 0.3+ 15: 1.2) minutes.

By adopting the method for determining the estimated delivery time length provided by the embodiment of the application, the weight corresponding to each delivery stage can be determined, and the stage time length corresponding to each delivery stage is subjected to weighted summation to obtain the estimated delivery time length corresponding to the whole delivery service. The method can reduce the risk of calculating deviation accumulation by directly summing the stage time length as the estimated delivery time length, thereby more accurately determining the estimated delivery time length corresponding to the completion of the delivery service.

The embodiment of the application provides a processing flow of the method for determining the estimated delivery duration as shown in fig. 5, which includes the following processing steps:

s301, acquiring duration estimation reference information.

S302, feature extraction is carried out on the time length estimation reference information based on the feature extraction model, and feature information corresponding to the time length estimation reference information is obtained.

S303, respectively inputting the characteristic information into the duration estimation models corresponding to the plurality of distribution stages to obtain the stage duration corresponding to each distribution stage.

And S304, determining the corrected time compensation duration based on the duration estimation reference information and the corrected time compensation calculation model.

The determination process of the corrected time-compensating duration may be as follows:

first, the server may determine the numerical parameters and the non-numerical parameters in the duration estimation reference information. Secondly, the server can perform discretization processing on the numerical parameter to obtain a discretization numerical parameter, and a second parameter vector is determined according to the discretization numerical parameter, the non-numerical parameter and the feature mapping model. And finally, the server inputs the second parameter vector into a correction time-supplementing calculation model, so that the correction time-supplementing duration corresponding to the distribution service can be determined. The second parameter vector is determined in step S102, and is not described herein again.

S305, determining the estimated delivery time length based on the corrected time-compensating time length and the stage time length corresponding to each delivery stage.

After determining the corrected time-compensating duration and the phase duration corresponding to each distribution phase, the server can sum the corrected time-compensating duration and the phase duration, so that the estimated distribution duration corresponding to the distribution service is determined. For example, the server determines that the period of time from the order placement of the user to the order pickup of the distributor is 5 minutes, the period of time from the order pickup of the distributor to the arrival of the distributor at the target merchant is 6 minutes, the period of time from the arrival of the distributor at the target merchant to the pickup of the goods by the distributor is 2 minutes, the period of time from the pickup of the distributor to the arrival of the distributor at the receiving place is 15 minutes, and the corrected replenishment period of time is 8 minutes, then the estimated delivery period of time for completing the delivery service is 36(5+6+2+15+8) minutes.

By adopting the method for determining the pre-estimated delivery time length provided by the embodiment of the application, the time length for correcting and supplementing the delivery service can be determined while each delivery stage in the whole delivery process is determined. The corrected time-compensating duration can be used for dealing with an unexpected situation occurring in the distribution process, for example, the duration consumed by a merchant for preparing goods is longer than the corresponding stage duration, the duration consumed by a salesman in the distribution process is longer than the corresponding stage duration due to serious road congestion, and the like. Therefore, certain buffering can be reserved for the distribution staff to finish the distribution business, the distribution pressure of the staff can be effectively relieved, and the distribution experience of the staff is improved.

The embodiment of the application provides a processing flow of the method for determining the estimated delivery duration as shown in fig. 6, which includes the following processing steps:

s401, obtaining duration estimation reference information.

S402, extracting the characteristics of the time length estimation reference information based on the characteristic extraction model to obtain the characteristic information corresponding to the time length estimation reference information.

And S403, inputting the characteristic information into the duration estimation models corresponding to the multiple distribution stages respectively to obtain the stage duration corresponding to each distribution stage.

S404, determining the weight corresponding to each distribution stage based on the duration estimation reference information and the weight calculation model.

In this step, the process of determining the weight corresponding to each delivery phase by the server is shown in step S204, and the second parameter vector input by the weight calculation model is the second parameter vector determined in step S102, which is not described herein again.

S405, determining the corrected time compensation duration based on the duration estimation reference information and the corrected time compensation calculation model.

In this step, the process of determining the corrected time duration by the server is shown in step S304, and the second parameter vector input by the corrected time duration calculation model is the second parameter vector determined in step S102, which is not described herein again.

S406, determining the estimated delivery time length based on the corrected time-compensating time length, the stage time length corresponding to each delivery stage and the weight.

After the server determines the corrected time-filling duration and the stage duration and the weight corresponding to each distribution stage, firstly, the weighted summation can be carried out on the duration of all the stages according to the weight corresponding to each distribution stage to obtain the sub-estimated distribution duration, and then the sub-estimated distribution duration is added with the corrected time-filling duration to obtain the estimated distribution duration corresponding to the distribution service. For example, the server determines that the period of time from the user placing the order to the distributor receiving the order is 5 minutes, the weight value is 0.5, the period of time from the distributor receiving the order to the distributor reaching the target merchant is 6 minutes, the weight value is 0.8, the period of time from the distributor reaching the target merchant to the distributor getting the goods is 2 minutes, the weight value is 0.3, the period of time from the distributor getting the goods to the distributor reaching the receiving place is 15 minutes, the weight value is 1.2, and the correction supplement time period is 8 minutes, so that the estimated distribution period of time for completing the distribution service is 29.58 (5: 0.5+ 6: 0.8+ 2: 0.3+ 15: 1.2+8) minutes.

By adopting the method for determining the estimated delivery time length, the weights of different delivery stages are considered, and the corrected time-compensating time length corresponding to the emergency is considered. Therefore, the stage duration corresponding to each distribution stage is accurately determined, a certain time buffer is reserved for the distributor, and the distribution experience of the salesmen is improved.

One or more abnormal judgment logics are stored in the server in advance and used for determining whether the estimated delivery time length is abnormal or not, for example, when the delivery distance is between 2.5 and 3 kilometers, the normal value range of the estimated delivery time length is more than or equal to 30 minutes and less than or equal to 50 minutes, and for example, when a merchant and a receiving place belong to the same administrative area, the estimated delivery time length cannot exceed 80 minutes, and the like. By adopting any method for determining the estimated delivery time length provided by the embodiment of the application, after the estimated delivery time length corresponding to the delivery service is determined, the server can also determine whether the estimated delivery time length is abnormal or not according to the prestored service logic. For example, when the service logic pre-stored in the server is that the delivery distance is between 2.5 and 3 kilometers, the normal value range of the estimated delivery duration is greater than or equal to 30 minutes and less than or equal to 50 minutes, if the estimated delivery duration corresponding to the delivery service with the delivery distance of 2.8 kilometers is 40 minutes, it is determined that the estimated delivery duration is not abnormal, and if the estimated delivery duration corresponding to the delivery service with the delivery distance of 2.8 kilometers is 25 minutes or 60 minutes, it is determined that the estimated delivery duration is abnormal.

If the estimated delivery time length is not abnormal, the server can send the estimated delivery time length to a terminal used by a user, and the estimated delivery time length is displayed to the user by the application program.

If the estimated delivery time length is abnormal, the server can re-determine the estimated delivery time length according to a pre-stored delivery time length threshold value. For example, when the service logic pre-stored in the server is that the delivery distance is between 2.5 and 3 kilometers, the estimated delivery time length is greater than or equal to 30 minutes and less than or equal to 50 minutes, if the estimated delivery time length corresponding to the delivery service with the delivery distance of 2.8 kilometers is 25 minutes, it is determined that the estimated delivery time length is abnormal, and the server will re-determine that the estimated delivery time length corresponding to the delivery service is 30 minutes. For another example, if the service logic is unchanged, if the estimated delivery time corresponding to the delivery service with the delivery distance of 2.8 km is 60 minutes, it is determined that the estimated delivery time is abnormal, and the server will re-determine that the estimated delivery time corresponding to the delivery service is 50 minutes. The server may send the re-determined estimated delivery duration to the terminal used by the user for display by the application to the user.

All the above optional technical solutions may be combined arbitrarily to form optional embodiments of the present application, and are not described herein again.

An embodiment of the present application provides a device for determining an estimated delivery duration, where the device may be a server in the foregoing embodiment, as shown in fig. 7, the device includes:

an obtaining module 710, configured to obtain duration estimation reference information, where the duration estimation reference information includes delivery order information, environment information corresponding to the delivery order information, and distributor distribution information;

the feature extraction module 720 is configured to perform feature extraction on the time length prediction reference information based on the feature extraction model to obtain feature information corresponding to the time length prediction reference information;

the stage duration estimation module 730 is configured to input the feature information into duration estimation models corresponding to the multiple distribution stages respectively to obtain a stage duration corresponding to each distribution stage;

the determining module 740 is configured to determine an estimated delivery duration based on a phase duration corresponding to each delivery phase.

the feature extraction module 720 is configured to:

In a possible implementation manner, the feature extraction module 720 is configured to:

the feature extraction module 720 is configured to:

In a possible implementation manner, the determining module 740 is configured to:

It should be noted that: in the device for determining the estimated delivery time provided in the above embodiment, when determining the estimated delivery time, only the division of the functional modules is used for illustration, and in practical applications, the function distribution may be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus for determining the estimated delivery time length and the method for determining the estimated delivery time length provided by the above embodiments belong to the same concept, and specific implementation processes thereof are detailed in the method embodiments and are not described herein again.

The computer device provided by the embodiment of the present application may be the server in the above embodiment. Fig. 8 is a schematic structural diagram of the computer device, the computer device 800 may generate a relatively large difference due to different configurations or performances, and may include one or more CPUs (processor) 810 and one or more memories 820, where at least one instruction is stored in the memory 820, and the at least one instruction is loaded and executed by the processor 810 to implement the methods provided by the above method embodiments. Certainly, the computer device may further have components such as a wired or wireless network interface, a keyboard, and an input/output interface, so as to perform input and output, and the computer device may further include other components for implementing the functions of the device, which is not described herein again.

In an exemplary embodiment, a computer-readable storage medium, such as a memory, is also provided that includes instructions executable by a processor in a terminal to perform the method of determining a pre-estimated delivery duration of the above embodiments. The computer readable storage medium may be non-transitory. For example, the computer-readable storage medium may be a ROM (Read-Only Memory), a RAM (Random Access Memory), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

The above description is only exemplary of the present application and should not be taken as limiting, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method of determining a pre-estimated delivery duration, the method comprising:

based on a feature extraction model, performing feature extraction on the duration estimation reference information to obtain feature information corresponding to the duration estimation reference information;

and determining the estimated delivery time length based on the stage time length corresponding to each delivery stage.

2. The method of claim 1, wherein the plurality of delivery phases include a phase from a user placing an order to a delivery person receiving an order, a phase from a delivery person receiving an order to a delivery person arriving at a delivery location, a phase from a delivery person arriving at a delivery location to a delivery person getting goods, a phase from a delivery person getting goods to a delivery person arriving at a receiving location.

3. The method of claim 1, wherein the delivery order information includes at least one of a consignor name, a goods name, a quantity of goods, a unit price of goods, an actual amount paid by a user, a delivery fee, location information of a delivery location, and location information of a receiving location;

the distributor distribution information includes at least one of the number of distributors whose distance from the delivery site does not exceed a distance threshold, and location information of distributors whose distance from the delivery site does not exceed a distance threshold.

4. The method of claim 1, wherein the feature extraction model comprises a first feature extraction module and a second feature extraction module, the first and second feature extraction modules being machine learning networks of different algorithms;

the characteristic extraction is carried out on the duration estimation reference information based on the characteristic extraction model, and the characteristic information corresponding to the duration estimation reference information is obtained, and the characteristic extraction comprises the following steps:

based on a first feature extraction module, performing feature extraction on the duration pre-estimation reference information to obtain first sub-feature information corresponding to the duration pre-estimation reference information;

based on a second feature extraction module, performing feature extraction on the duration pre-estimation reference information to obtain second sub-feature information corresponding to the duration pre-estimation reference information;

5. The method according to claim 4, wherein the extracting the feature of the duration estimation reference information based on the first feature extraction module to obtain the first sub-feature information corresponding to the duration estimation reference information comprises:

inputting the first parameter vector into the first feature extraction module to obtain first sub-feature information corresponding to the duration pre-estimation reference information;

the second feature extraction module is used for extracting features of the duration pre-estimation reference information to obtain second sub-feature information corresponding to the duration pre-estimation reference information, and the second sub-feature information comprises:

carrying out split-point discretization on the numerical parameter in the duration estimation reference information to obtain a discretization numerical parameter;

determining a second parameter vector based on the discretization numerical parameter and a non-numerical parameter in the duration pre-estimation reference information;

and inputting the second parameter vector into the second feature extraction module to obtain second sub-feature information corresponding to the duration pre-estimation reference information.

6. The method of claim 1, wherein determining the estimated delivery duration based on the phase duration corresponding to each delivery phase comprises:

determining the weight corresponding to each distribution stage based on the duration estimation reference information and a weight calculation model;

and determining the estimated delivery time length based on the corresponding stage time length and weight of each delivery stage.

7. The method of claim 6, wherein the determining the weight corresponding to each delivery phase based on the duration estimation reference information and the weight calculation model comprises:

and inputting the second parameter vector into a weight calculation model to obtain the weight corresponding to each distribution stage.

8. The method of claim 1, wherein determining the estimated delivery duration based on the phase duration corresponding to each delivery phase comprises:

determining a correction time-supplementing duration based on the duration pre-estimation reference information and a correction time-supplementing calculation model;

and determining the estimated delivery duration based on the corrected time-compensating duration and the stage duration corresponding to each delivery stage.

9. The method of claim 8, wherein determining a modified time-filling duration based on the duration estimation reference information and a modified time-filling calculation model comprises:

10. An apparatus for determining an estimated delivery duration, the apparatus comprising:

the characteristic extraction module is used for extracting the characteristics of the duration estimation reference information based on a characteristic extraction model to obtain characteristic information corresponding to the duration estimation reference information;

the stage duration estimation module is used for respectively inputting the characteristic information into duration estimation models corresponding to a plurality of distribution stages to obtain stage duration corresponding to each distribution stage;

11. A computer device comprising a processor and a memory, the memory having stored therein at least one instruction that is loaded into and executed by the processor to perform operations performed by the method of determining an estimated delivery duration of any of claims 1 to 9.

12. A computer-readable storage medium having stored therein at least one instruction which is loaded into and executed by a processor to perform operations performed by the method of determining a pre-estimated delivery duration of any of claims 1 to 9.

13. A computer program product, characterized in that the computer program product comprises computer program code which, when executed by a computer arrangement, executes the method of determining a pre-estimated delivery duration of any of the preceding claims 1 to 9.