Model training method, route freight total volume generation method, device and equipment

Abstract

The embodiment of the invention discloses a model training method, a route freight total volume generation method, a device and equipment. A specific embodiment of the method comprises the following steps: acquiring an off-line route freight data set and a real-time route freight data set; performing data correction on the offline route freight data set and the real-time route freight data set to generate a corrected offline route freight data set and a corrected real-time route freight data set; generating an offline feature data set and a real-time feature data set according to the corrected offline route freight data set and the corrected real-time route freight data set; and according to the offline feature data set and the real-time feature data set, performing model training on an initial real-time route freight total volume generation model to generate a real-time route freight total volume generation model. The implementation mode is related to artificial intelligence, and a more accurate real-time route freight total volume can be generated by utilizing the real-time route freight total volume generation model.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a model training method, a method, a device, and an apparatus for generating the total freight volume of a route. Background Art

[0002] With the rapid development of the express delivery industry, logistics and transportation costs are also gradually increasing. The optimal transportation of goods along a route has become a key concern. Forecasting cargo volume along a target route is typically done through manual empirical methods to predict the volume of cargo along the target route during a target future time period.

[0003] However, the inventors have discovered that when the above method is adopted, the following technical problems often occur:

[0004] The quality of historical cargo volume data for the target route is poor, which not only affects manual judgment of the cargo volume in the target future time period, but also the manual judgment itself has problems with accuracy and efficiency.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0006] The content of this disclosure is used to briefly introduce concepts that will be described in detail in the detailed description section below. The content of this disclosure is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0007] Some embodiments of the present disclosure propose a model training method, a method, apparatus, and device for generating the total freight volume of a route to solve the technical problems mentioned in the above background technology section.

[0008] In a first aspect, some embodiments of the present disclosure provide a model training method, including: obtaining an offline route freight dataset and a real-time route freight dataset; performing data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; generating an offline feature dataset and a real-time feature dataset based on the corrected offline route freight dataset and the corrected real-time route freight dataset; performing model training on an initial real-time route freight total volume generation model based on the offline feature dataset and the real-time feature dataset to generate a real-time route freight total volume generation model.

[0009] Optionally, the offline route freight data in the above-mentioned offline route freight data set includes: historical actual total shipment volume and historical actual shipment vehicle information; and the above-mentioned data correction of the above-mentioned offline route freight data set and the above-mentioned real-time route freight data set to generate a corrected offline route freight data set and a corrected real-time route freight data set includes: for each offline route freight data in the above-mentioned offline route freight data set, performing the following first correction operation: generating the historical actual shipment vehicle total volume according to the historical actual shipment vehicle information corresponding to the above-mentioned offline route freight data; in response to determining that the historical actual shipment total volume corresponding to the above-mentioned offline route freight data is greater than the above-mentioned historical actual shipment vehicle total volume, correcting the historical actual shipment total volume corresponding to the above-mentioned offline route freight data to the above-mentioned historical actual shipment vehicle total volume to obtain the corrected offline route freight data.

[0010] Optionally, the real-time route freight data in the above-mentioned real-time route freight data set includes: the estimated total volume of route freight for the target departure time; and the above-mentioned data correction of the above-mentioned offline route freight data set and the above-mentioned real-time route freight data set to generate a corrected offline route freight data set and a corrected real-time route freight data set, including: for each real-time route freight data in the above-mentioned real-time route freight data set, performing the following second correction operation: determining the corrected offline route freight data in the same period as the above-mentioned real-time route freight data as the target corrected offline route freight data; in response to the historical actual shipment total volume corresponding to the above-mentioned target corrected offline route freight data being less than the route freight estimated total volume corresponding to the above-mentioned real-time route freight data, correcting the route freight estimated total volume corresponding to the above-mentioned real-time route freight data to the historical actual shipment total volume corresponding to the above-mentioned corrected offline route freight data.

[0011] Optionally, in response to the above-mentioned target correction, the historical actual total shipment volume corresponding to the offline route freight data is less than the route freight estimated total volume corresponding to the above-mentioned real-time route freight data, and the route freight estimated total volume corresponding to the above-mentioned real-time route freight data is corrected to the historical actual total shipment volume corresponding to the above-mentioned corrected offline route freight data. The above-mentioned method also includes: in response to determining that the route freight estimated total volume corresponding to the above-mentioned real-time route freight data is a negative number, correcting the route freight estimated total volume corresponding to the above-mentioned real-time route freight data to the target value.

[0012] Optionally, the above-mentioned generation of the offline feature dataset and the real-time feature dataset based on the above-mentioned corrected offline route freight dataset and the above-mentioned corrected real-time route freight dataset includes: obtaining an offline feature information set; extracting an offline feature dataset corresponding to the above-mentioned offline feature information set from the above-mentioned corrected offline route freight dataset; for each corrected real-time route freight data in the above-mentioned corrected real-time route freight dataset, performing the following generation steps: determining a subset of corrected real-time route freight data at the same time point as the above-mentioned corrected real-time route freight data; determining the quantile of the data size corresponding to the above-mentioned corrected real-time route freight data in the above-mentioned corrected real-time route freight data subset; determining the corrected offline route freight data in the above-mentioned corrected offline route freight dataset whose corresponding data size quantile is the above-mentioned quantile as the quantile value; and determining the fusion information of the above-mentioned quantile and the above-mentioned quantile value as the real-time feature data for the above-mentioned corrected real-time route freight data.

[0013] Optionally, the above-mentioned corrected real-time route freight data includes: the real-time route freight order quantity, and the above-mentioned quantile includes the first quantile; and the above-mentioned determination of the quantile of the data size corresponding to the above-mentioned corrected real-time route freight data in the above-mentioned corrected real-time route freight data subset includes: determining the quantile of the order quantity size of the real-time route freight order quantity corresponding to the above-mentioned corrected real-time route freight data in the quantile subset as the above-mentioned first quantile, wherein the above-mentioned quantile subset is the real-time route freight order subset corresponding to the above-mentioned corrected real-time route freight data subset.

[0014] Optionally, the above-mentioned percentile value includes a first percentile value; and the above-mentioned determination of the corrected offline route freight data set whose corresponding data size percentile is the above-mentioned percentile as the percentile value includes: determining the corrected offline route freight data set whose corresponding data size percentile is the above-mentioned first percentile as the first percentile value.

[0015] Optionally, the above-mentioned corrected real-time route freight data includes: the estimated total volume of route freight, and the above-mentioned quantile includes a second quantile; and the above-mentioned determination of the quantile of the data size corresponding to the above-mentioned corrected real-time route freight data in the above-mentioned corrected real-time route freight data subset includes: determining the quantile of the volume size of the estimated total volume of route freight corresponding to the above-mentioned corrected real-time route freight data in the volume subset as the above-mentioned second quantile, wherein the above-mentioned volume subset is the estimated total volume subset of route freight corresponding to the above-mentioned corrected real-time route freight data subset.

[0016] Optionally, the above-mentioned quantile value includes a second quantile value; and the above-mentioned determination of the corrected offline route freight data set whose corresponding data size quantile is the above-mentioned quantile as the corrected offline route freight data as the quantile value includes: determining the corrected offline route freight data set whose corresponding data size quantile is the above-mentioned second quantile as the second quantile value.

[0017] In a second aspect, some embodiments of the present disclosure provide a model training device, including: a first acquisition unit, configured to acquire an offline route freight dataset and a real-time route freight dataset; a correction unit, configured to perform data correction on the above-mentioned offline route freight dataset and the above-mentioned real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; a generation unit, configured to generate an offline feature dataset and a real-time feature dataset based on the above-mentioned corrected offline route freight dataset and the above-mentioned corrected real-time route freight dataset; a training unit, configured to perform model training on an initial real-time route freight total volume generation model based on the above-mentioned offline feature dataset and the above-mentioned real-time feature dataset to generate a real-time route freight total volume generation model.

[0018] Optionally, the offline route freight data in the above-mentioned offline route freight data set includes: historical actual shipment total volume and historical actual shipment vehicle information; and the correction unit can be configured to: for each offline route freight data in the above-mentioned offline route freight data set, perform the following first correction operation: generate the historical actual shipment vehicle total volume according to the historical actual shipment vehicle information corresponding to the above-mentioned offline route freight data; in response to determining that the historical actual shipment total volume corresponding to the above-mentioned offline route freight data is greater than the above-mentioned historical actual shipment vehicle total volume, correct the historical actual shipment total volume corresponding to the above-mentioned offline route freight data to the above-mentioned historical actual shipment vehicle total volume, and obtain the corrected offline route freight data.

[0019] Optionally, the real-time route freight data in the above-mentioned real-time route freight data set includes: an estimated total volume of route freight for a target departure time; and the correction unit can be configured to: for each real-time route freight data in the above-mentioned real-time route freight data set, perform the following second correction operation: determine the corrected offline route freight data in the same period as the above-mentioned real-time route freight data as the target corrected offline route freight data; in response to the historical actual shipment total volume corresponding to the above-mentioned target corrected offline route freight data being smaller than the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data, correct the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data to the historical actual shipment total volume corresponding to the above-mentioned corrected offline route freight data.

[0020] Optionally, the correction unit may be configured to: in response to determining that the estimated total volume of route freight corresponding to the real-time route freight data is a negative number, correct the estimated total volume of route freight corresponding to the real-time route freight data to a target value.

[0021] Optionally, the generation unit can be configured to: obtain an offline feature information set; extract an offline feature data set corresponding to the above-mentioned offline feature information set from the above-mentioned corrected offline route freight data set; for each corrected real-time route freight data in the above-mentioned corrected real-time route freight data set, perform the following generation steps: determine a subset of corrected real-time route freight data at the same time point as the above-mentioned corrected real-time route freight data; determine the quantile of the data size corresponding to the above-mentioned corrected real-time route freight data in the above-mentioned corrected real-time route freight data subset; determine the corrected offline route freight data in the above-mentioned corrected offline route freight data set whose corresponding data size quantile is the above-mentioned quantile as the quantile value; and determine the fusion information of the above-mentioned quantile and the above-mentioned quantile value as the real-time feature data for the above-mentioned corrected real-time route freight data.

[0022] Optionally, the above-mentioned corrected real-time route freight data includes: the real-time route freight order quantity, and the above-mentioned quantile includes a first quantile; and the generation unit can be configured to: determine the quantile of the order quantity size of the real-time route freight order quantity corresponding to the above-mentioned corrected real-time route freight data in the quantile subset as the above-mentioned first quantile, wherein the above-mentioned quantile subset is the real-time route freight order subset corresponding to the above-mentioned corrected real-time route freight data subset.

[0023] Optionally, the quantile value includes a first quantile value; and the generating unit may be configured to: determine the corrected offline route freight data set whose corresponding data size quantile in the corrected offline route freight data set is the first quantile point as the first quantile value.

[0024] Optionally, the above-mentioned corrected real-time route freight data includes: the estimated total volume of route freight, and the above-mentioned quantile includes a second quantile; and the generation unit can be configured to: determine the quantile of the volume size of the estimated total volume of route freight corresponding to the above-mentioned corrected real-time route freight data in the volume subset as the above-mentioned second quantile, wherein the above-mentioned volume subset is the estimated total volume subset of route freight corresponding to the above-mentioned corrected real-time route freight data subset.

[0025] Optionally, the quantile value includes a second quantile value; and the generating unit may be configured to: determine the corrected offline route freight data set whose corresponding data size quantile in the corrected offline route freight data set is the second quantile point as the second quantile value.

[0026] In a third aspect, some embodiments of the present disclosure provide a method for generating a total route freight volume, comprising: obtaining an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset; inputting the above target offline feature dataset and the above target real-time feature dataset into a pre-trained real-time route freight total volume generation model to generate a predicted total route freight volume for the above future prediction time period, wherein the above real-time route freight total volume generation model is generated based on the model training method corresponding to the first aspect; performing volume correction on the above predicted total route freight volume based on a pre-generated range interval of the route freight total volume to generate a corrected route freight total volume.

[0027] Optionally, the method further includes: generating vehicle dispatching information for the target route based on the total freight volume of the corrected route; and executing dispatching processing of relevant vehicles based on the vehicle dispatching information.

[0028] In a fourth aspect, some embodiments of the present disclosure provide a route freight total volume generation device, comprising: a second acquisition unit, configured to acquire an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset; a second input unit, configured to input the above-mentioned target offline feature dataset and the above-mentioned target real-time feature dataset into a pre-trained real-time route freight total volume generation model to generate a route freight prediction total volume for the above-mentioned future prediction time period, wherein the above-mentioned real-time route freight total volume generation model is generated based on the model training method corresponding to the first aspect; a correction unit, configured to perform volume correction on the above-mentioned route freight prediction total volume based on a pre-generated route freight total volume range interval to generate a corrected route freight total volume.

[0029] Optionally, the device further includes: generating vehicle dispatching information for the target route according to the total freight volume of the corrected route; and executing dispatching processing of relevant vehicles according to the vehicle dispatching information.

[0030] In a fifth aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by one or more processors, the one or more processors implement the method described in any one of the implementation methods in the first and third aspects.

[0031] In a sixth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any one of the implementation modes of the first and third aspects is implemented.

[0032] In a seventh aspect, some embodiments of the present disclosure provide a computer program product, including a computer program, which, when executed by a processor, implements the method described in any one of the implementation modes of the first and third aspects above.

[0033] The above-described embodiments of the present disclosure have the following beneficial effects: The model training methods of some embodiments of the present disclosure can utilize a real-time route freight volume generation model to generate a more accurate real-time route freight volume. Specifically, the reason for the inaccurate cargo volume situation is that the quality of historical freight volume data for the target route is poor. This not only affects manual judgment of the cargo volume situation for the target future time period, but also leads to inaccurate and inefficient manual judgment itself. Based on this, the model training methods of some embodiments of the present disclosure first obtain an offline route freight dataset and a real-time route freight dataset to subsequently generate an offline feature dataset and a real-time feature dataset. Then, data correction is performed on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset. Here, by performing data correction on the offline route freight dataset and the real-time route freight dataset, data quality is improved, the accuracy of the subsequent offline feature dataset and the real-time feature dataset is guaranteed, and the problem of inaccurate training of the initial real-time route freight volume generation model is avoided. Furthermore, an offline feature dataset and a real-time feature dataset are generated based on the corrected offline route freight dataset and the corrected real-time route freight dataset. Here, the generated offline feature dataset and real-time feature dataset can characterize various aspects of route freight characteristics, so that in the subsequent training process of the initial real-time route freight total volume generation model, the initial real-time route freight total volume generation model can learn more feature content. Finally, based on the above-mentioned offline feature dataset and the above-mentioned real-time feature dataset, the initial real-time route freight total volume generation model is trained to generate a real-time route freight total volume generation model. In summary, by performing data correction on the above-mentioned offline route freight dataset and the above-mentioned real-time route freight dataset, the data quality can be improved. In this way, the offline feature dataset and the real-time feature dataset can be accurately generated to obtain a real-time route freight total volume generation model that generates a more accurate real-time route freight total volume. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0035] Figure 1is a schematic diagram of an application scenario of the model training method according to some embodiments of the present disclosure;

[0036] Figure 2 is a flowchart of some embodiments of the model training method according to the present disclosure;

[0037] Figure 3 is a flowchart of other embodiments of the model training method according to the present disclosure;

[0038] Figure 4 is a flow chart of some embodiments of a method for generating a total volume of freight for a route according to the present disclosure;

[0039] Figure 5 is a schematic structural diagram of some embodiments of the model training device according to the present disclosure;

[0040] Figure 6 is a schematic structural diagram of some embodiments of a device for generating a total volume of freight on a route according to the present disclosure;

[0041] Figure 7 It is a structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0042] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0043] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other.

[0044] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0045] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, they should be understood as "one or more".

[0046] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0047] Before performing the collection, storage, and use of the data involved in this disclosure (such as offline route freight data and real-time route freight data), the relevant organizations or individuals must fulfill their obligations, including data security impact assessments, notification obligations to data subjects, and prior authorization and consent from the subject of personal information.

[0048] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0049] Figure 1 It is a schematic diagram of an application scenario of the model training method according to some embodiments of the present disclosure.

[0050] exist Figure 1 In the application scenario, first, the electronic device 101 can obtain an offline route freight dataset 102 and a real-time route freight dataset 103. Then, the electronic device 101 can perform data correction on the offline route freight dataset 102 and the real-time route freight dataset 103 to generate a corrected offline route freight dataset 106 and a corrected real-time route freight dataset 107. Next, the electronic device 101 can generate an offline feature dataset 106 and a real-time feature dataset 107 based on the corrected offline route freight dataset 106 and the corrected real-time route freight dataset 107. Finally, based on the offline feature dataset 106 and the real-time feature dataset 107, the initial real-time route freight total volume generation model 108 is trained to generate a real-time route freight total volume generation model 109.

[0051] It should be noted that the electronic device 101 can be hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or it can be implemented as a single server or a single terminal device. When the electronic device is embodied as software, it can be installed in the hardware devices listed above. It can be implemented as multiple software or software modules for providing distributed services, for example, or it can be implemented as a single software or software module. No specific limitation is made here.

[0052] It should be understood that Figure 1 The number of electronic devices in the embodiment is merely illustrative. Any number of electronic devices may be provided according to implementation requirements.

[0053] Continue to refer Figure 2 , shows a process 200 of some embodiments of the model training method according to the present disclosure. The model training method includes the following steps:

[0054] Step 201: Obtain an offline route freight dataset and a real-time route freight dataset.

[0055] In some embodiments, the execution entity of the above model training method (for example Figure 1 The electronic device 101 shown can obtain an offline route freight dataset and a real-time route freight dataset via a wired or wireless connection. The offline route freight dataset and the real-time route freight dataset can be offline datasets and real-time datasets corresponding to at least one route. One of the at least one route can be a freight transportable route. The offline route freight data in the offline route freight dataset can be historical route freight data within a first predetermined time granularity. In practice, the first predetermined time granularity can be a daily granularity, for example, one day. The historical route freight data can be freight data corresponding to the route. In practice, the historical route freight data can include, but is not limited to, at least one of the following: number of shipments for the route, number of shipments for the route, and number of shipments by vehicle. For example, for route A, the corresponding offline route freight dataset can include freight data corresponding to November 1, November 2, November 3, ..., and November 30. The real-time route freight data in the real-time route freight dataset can be historical route freight data within a second predetermined time granularity. The time granularity corresponding to the second predetermined time granularity is smaller than the time granularity corresponding to the first predetermined time granularity. In practice, the second predetermined time granularity may be hourly granularity, for example, one hour. Specifically, the real-time route freight data in the real-time route freight data set may be freight data within a pre-target time period before the departure time. The departure time may be the departure time of the truck to be driven corresponding to the route. Specifically, for a certain route, there are trucks to be driven every day to transport goods, that is, each route has a corresponding truck departure time for the truck to be driven. For example, the pre-target time period may be within the previous 1 hour to 48 hours. For example, for November 1, the corresponding departure time on November 1 is "18:00", and the corresponding at least one real-time route freight data may include: "real-time route freight data corresponding to 17:00", "real-time route freight data corresponding to 16:00", "real-time route freight data corresponding to 15:00", and "real-time route freight data corresponding to 14:00".

[0056] Step 202 : Correcting the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset.

[0057] In some embodiments, the execution entity may perform data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset.

[0058] As an example, the execution entity may fill in the empty data in the offline route freight dataset and the real-time route freight dataset to generate a revised offline route freight dataset and a revised real-time route freight dataset.

[0059] In some optional implementations of some embodiments, the offline route freight data in the offline route freight dataset includes: historical actual total shipment volume and historical actual shipment vehicle information. The historical actual shipment volume may be the total volume of freight shipped for the route within a first predetermined time granularity. The historical actual shipment vehicle information may be the information of trucks shipped for the route within the first predetermined time granularity. In practice, the truck vehicle information may include, but is not limited to, at least one of the following: truck license plate number and truck cargo capacity.

[0060] Optionally, the data correction of the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset may include the following steps:

[0061] For each offline route freight data in the above offline route freight dataset, perform the following first correction operation:

[0062] Sub-step 1: Generate the total volume of historically actual dispatched vehicles based on the historically actual dispatched vehicles information corresponding to the offline route freight data. The total volume of historically actual dispatched vehicles may be the total volume of cargo that can be loaded by vehicles dispatched within a first predetermined time granularity for the route.

[0063] As an example, first, the shipping vehicle information corresponding to the offline route freight data is determined to obtain at least one shipping vehicle information. Then, the cargo loadable volume of the truck corresponding to each of the at least one shipping vehicle information is determined to obtain at least one cargo loadable volume. Finally, the cargo loadable volumes of each truck corresponding to the at least one cargo loadable volume are added together to obtain the total cargo loadable volume of the vehicles.

[0064] Sub-step 2, in response to determining that the historical actual total shipment volume corresponding to the above-mentioned offline route freight data is greater than the above-mentioned historical actual total shipment vehicle volume, the historical actual total shipment volume corresponding to the above-mentioned offline route freight data is corrected to the above-mentioned historical actual total shipment vehicle volume to obtain the corrected offline route freight data.

[0065] In some optional implementations of some embodiments, the real-time route freight data in the above-mentioned real-time route freight dataset includes: an estimated total route freight volume for a target departure time. The estimated total route freight volume may be the total route freight volume for the day corresponding to the target departure time and at least one day thereafter. For example, the current time is 3:00 PM on June 1st, and the target departure time is 6:00 PM on June 1st. The estimated total route freight volume may be the estimated total route freight volume for "June 1st - June 2nd." For another example, the current time is 3:00 PM on June 1st, and the target departure time is 6:00 PM on June 2nd. The estimated total route freight volume may be the estimated total route freight volume for "June 2nd - June 3rd." In practice, the estimated total route freight volume may be the estimated total route freight volume for the target departure time, based on the number of orders placed at the current time. Specifically, the estimated total route freight volume may be estimated using a Transformer model.

[0066] Optionally, the data correction of the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset includes:

[0067] For each real-time route freight data in the real-time route freight data set, perform the following second correction operation:

[0068] Sub-step 1: determining the corrected offline route freight data that is in the same period as the real-time route freight data as the target corrected offline route freight data. The same period may be data on the same day as the real-time route freight data and the corrected offline route freight data.

[0069] Sub-step 2: In response to the above-mentioned target-corrected offline route freight data having a smaller historical actual total shipment volume than the above-mentioned real-time route freight data having a predicted total route freight volume, the execution entity may correct the above-mentioned real-time route freight data having a predicted total route freight volume to the above-mentioned corrected offline route freight data having a predicted total historical actual shipment volume.

[0070] Optionally, in response to the above target-corrected offline route freight data indicating that the historical actual total shipment volume corresponding to the target-corrected offline route freight data is smaller than the estimated total route freight volume corresponding to the real-time route freight data, after correcting the estimated total route freight volume corresponding to the real-time route freight data to the historical actual total shipment volume corresponding to the corrected offline route freight data, the method further includes:

[0071] In response to determining that the estimated total volume of freight for the route corresponding to the real-time freight data is a negative number, the execution entity may correct the estimated total volume of freight for the route corresponding to the real-time freight data to a target value. The target value may be a preset value. For example, the target value may be "0."

[0072] Step 203 : generating an offline feature dataset and a real-time feature dataset based on the revised offline route freight dataset and the revised real-time route freight dataset.

[0073] In some embodiments, the execution entity may generate an offline feature dataset and a real-time feature dataset based on the revised offline route freight dataset and the revised real-time route freight dataset. The offline feature data in the offline feature dataset may be data corresponding to the offline feature. The offline feature may be a pre-set feature. Similarly, the real-time feature data in the real-time feature dataset may be data corresponding to the real-time feature. The real-time feature may be a pre-set feature. Optionally, the offline feature may be a feature associated with the long-term feature. The real-time feature may be a feature associated with the short-term feature. For example, the offline feature may include: the variance of the total volume transformation of route freight over a long period of time, and the combination feature of vehicle models of route freight. The short-term feature may include: the order quantity feature in a short period of time, the transformation feature of the total volume of route freight in a short period of time,

[0074] As an example, the execution entity may first determine an offline feature set and a real-time feature set. Then, the execution entity may extract an offline feature set and a real-time feature set associated with the offline feature set and the real-time feature set from the revised offline route freight data set and the revised real-time route freight data set, respectively.

[0075] Step 204 : Based on the offline feature dataset and the real-time feature dataset, the initial real-time route freight volume generation model is trained to generate a real-time route freight volume generation model.

[0076] In some embodiments, the execution entity may perform model training on the initial real-time route freight volume generation model based on the offline feature dataset and the real-time feature dataset to generate a real-time route freight volume generation model. The initial real-time route freight volume generation model may be a real-time route freight volume generation model for which training has not yet been completed. The real-time route freight volume generation model may be a model that generates the total route freight volume corresponding to a route within a predetermined time period in the future. In practice, the real-time route freight volume generation model may be a machine learning model. For example, the real-time route freight volume generation model may be an XGBoost model.

[0077] As an example, the execution entity may use the fused feature dataset of the offline feature dataset and the real-time feature dataset as a training dataset to perform model training on the initial real-time route freight volume generation model to generate a real-time route freight volume generation model.

[0078] The above-described embodiments of the present disclosure have the following beneficial effects: The model training methods of some embodiments of the present disclosure can utilize a real-time route freight volume generation model to generate a more accurate real-time route freight volume. Specifically, the reason for the inaccurate cargo volume situation is that the quality of historical freight volume data for the target route is poor. This not only affects manual judgment of the cargo volume situation for the target future time period, but also leads to inaccurate and inefficient manual judgment itself. Based on this, the model training methods of some embodiments of the present disclosure first obtain an offline route freight dataset and a real-time route freight dataset to subsequently generate an offline feature dataset and a real-time feature dataset. Then, data correction is performed on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset. Here, by performing data correction on the offline route freight dataset and the real-time route freight dataset, data quality is improved, the accuracy of the subsequent offline feature dataset and the real-time feature dataset is guaranteed, and the problem of inaccurate training of the initial real-time route freight volume generation model is avoided. Furthermore, an offline feature dataset and a real-time feature dataset are generated based on the corrected offline route freight dataset and the corrected real-time route freight dataset. Here, the generated offline feature dataset and real-time feature dataset can characterize various aspects of route freight characteristics, so that in the subsequent training process of the initial real-time route freight total volume generation model, the initial real-time route freight total volume generation model can learn more feature content. Finally, based on the above-mentioned offline feature dataset and the above-mentioned real-time feature dataset, the initial real-time route freight total volume generation model is trained to generate a real-time route freight total volume generation model. In summary, by performing data correction on the above-mentioned offline route freight dataset and the above-mentioned real-time route freight dataset, the data quality can be improved. In this way, the offline feature dataset and the real-time feature dataset can be accurately generated to obtain a real-time route freight total volume generation model that generates a more accurate real-time route freight total volume.

[0079] Further references Figure 3 , shows a process 300 of another embodiment of the model training method according to the present disclosure. The model training method includes the following steps:

[0080] Step 301: Obtain an offline route freight data set and a real-time route freight data set.

[0081] Step 302 : Correcting the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset.

[0082] Step 303: Obtain an offline feature information set.

[0083] In some embodiments, the execution entity (e.g. Figure 1 The electronic device 101 shown) can obtain an offline feature information set. The offline feature information can represent the feature identity information of the offline feature. For example, the offline feature information can be the noun information of the offline feature. For example, the offline feature information set includes: time feature information and statistical feature information. Specifically, the time feature information can include: derived features of the predicted date. For example, the derived features of the predicted date can include: weekday features, holiday features, week features, and month features. The statistical feature information can be feature information constructed based on a fixed sliding window method. For example, the fixed sliding window is 7 days, and features such as the maximum value, minimum value, and average value of the time window are extracted, and feature construction can be performed from two dimensions: the historical number of route freight orders and the total volume of historical route freight.

[0084] Step 304: extracting an offline feature data set corresponding to the offline feature information set from the corrected offline route freight data set.

[0085] In some embodiments, the execution entity may extract an offline feature data set corresponding to the offline feature information set from the corrected offline route freight data set.

[0086] Step 305: For each corrected real-time freight route data in the corrected real-time freight route data set, perform the following generation steps:

[0087] Step 3051: Determine a subset of the revised real-time route freight data at the same time point as the revised real-time route freight data.

[0088] In some embodiments, the execution entity may determine a subset of revised real-time route freight data that coincides with the revised real-time route freight data. The same time period may be time points on different days with the same time interval from departure. For example, the revised real-time route freight data corresponds to "13:00 on June 11th," with six hours remaining until departure. The revised real-time route freight data subset may include revised real-time route freight data corresponding to "June 10th, six hours until departure," revised real-time route freight data corresponding to "June 9th, six hours until departure," revised real-time route freight data corresponding to "June 8th, six hours until departure," revised real-time route freight data corresponding to "June 7th, six hours until departure," and revised real-time route freight data corresponding to "June 6th, six hours until departure."

[0089] Step 3052: Determine the quantile of the data size corresponding to the corrected real-time route freight data in the corrected real-time route freight data subset.

[0090] In some embodiments, the execution entity may determine a quantile of the data size corresponding to the corrected real-time route freight data within the corrected real-time route freight data subset. The quantile represents the data size position of the corrected real-time route freight data within the corrected real-time route freight data subset.

[0091] As an example, the execution entity may first determine the freight cost corresponding to the corrected real-time route freight data and a freight cost subset corresponding to the corrected real-time route freight data subset. Then, the freight cost corresponding to the corrected real-time route freight data is determined in terms of cost magnitude within the freight cost subset to obtain a quantile.

[0092] In some optional implementations of some embodiments, the revised real-time route freight data includes: a real-time route freight order volume. The quantiles include a first quantile. The real-time route freight order volume may be the number of freight orders shipped for the route corresponding to the revised real-time route freight data within a corresponding time period. The first quantile may represent the position of the freight order volume corresponding to the revised real-time route freight data within a subset of the revised real-time route freight data.

[0093] Optionally, the execution entity may determine a quantile of the order quantity of the real-time route freight order quantity corresponding to the revised real-time route freight data within a sub-quantile subset as the first quantile. The sub-quantile subset is the real-time route freight order quantity subset corresponding to the revised real-time route freight data subset. There is a one-to-one correspondence between the revised real-time route freight data in the revised real-time route freight data subset and the real-time route freight order quantity in the real-time route freight order quantity subset.

[0094] In some optional implementations of some embodiments, the corrected real-time route freight data includes an estimated total volume of route freight. The quantile includes a second quantile. The second quantile represents the volume position of the estimated total volume of route freight corresponding to the corrected real-time route freight data within the corrected real-time route freight data subset.

[0095] Optionally, the execution entity may determine a quantile of the estimated total route freight volume corresponding to the corrected real-time route freight data within a volume subset as the second quantile, wherein the volume subset is a subset of the estimated total route freight volume corresponding to the corrected real-time route freight data subset. There is a one-to-one correspondence between the corrected real-time route freight data in the corrected real-time route freight data subset and the estimated total route freight volume in the estimated total route freight volume subset.

[0096] Step 3053: Determine the corrected offline route freight data set whose corresponding data size quantile is the above quantile in the corrected offline route freight data set as the quantile value.

[0097] In some embodiments, the execution entity may determine the corrected offline route freight data set having a corresponding data size quantile as the quantile point in the corrected offline route freight data set as the quantile value.

[0098] As an example, for a quantile of 0.5, the corrected offline route freight data set is sorted by freight cost size to obtain a corrected offline route freight data sequence. Then, the corrected offline route freight data with a freight cost size quantile of 0.5 is determined as the quantile value.

[0099] In some optional implementations of some embodiments, the above-mentioned quantile value includes a first quantile value.

[0100] Optionally, the execution entity may determine the corrected offline route freight data whose corresponding data size quantile in the corrected offline route freight data set is the first quantile as the first quantile value.

[0101] In some optional implementations of some embodiments, the above-mentioned quantile value includes a second quantile value.

[0102] Optionally, the execution entity may determine the corrected offline route freight data whose corresponding data size quantile in the corrected offline route freight data set is the second quantile as the second quantile value.

[0103] In step 3054, the fusion information of the quantile point and the quantile value is determined as the real-time feature data for the corrected real-time route freight data.

[0104] In some embodiments, the execution entity may determine the fusion information of the quantile point and the quantile value as the real-time feature data for the corrected real-time route freight data.

[0105] Step 306 : Based on the offline feature dataset and the real-time feature dataset, the initial real-time route freight volume generation model is trained to generate a real-time route freight volume generation model.

[0106] In some embodiments, the specific implementation of steps 301-302 and 306 and the technical effects thereof can be referred to in Figure 2 Steps 201 - 202 and 204 in the corresponding embodiment will not be repeated here.

[0107] from Figure 3 It can be seen that Figure 2 Compared with the description of some corresponding embodiments, Figure 3 In the corresponding process 300 of the model training method in some embodiments, by determining a subset of the corrected real-time route freight data at the same time point as the above-mentioned corrected real-time route freight data, the quantiles and quantile values ​​are subsequently determined, which can more accurately reflect the characteristic content of the real-time features, making subsequent model training more accurate and avoiding the problem of unstable output of the model.

[0108] Continue to refer Figure 4 , shows a process 400 of some embodiments of the method for generating the total freight volume of a route according to the present disclosure. The method for generating the total freight volume of a route includes the following steps:

[0109] Step 401 : Obtain an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset.

[0110] In some embodiments, the execution entity (e.g., an electronic device) of the above-mentioned route total freight volume generation method can obtain an offline feature dataset and a real-time feature dataset for the target route through a wired connection or a wireless connection, as a target offline feature dataset and a target real-time feature dataset. The target route can be a route for which total freight volume prediction is to be performed. The target offline feature dataset can be an offline feature dataset for the first predetermined number of days before the departure time. The target real-time feature dataset can be a real-time feature dataset for the second predetermined number of hours before the departure time. For example, the second predetermined number of hours can be 48. The first predetermined number of days can be 30 days.

[0111] Step 402 : Input the target offline feature dataset and the target real-time feature dataset into a pre-trained real-time route freight volume generation model to generate a predicted route freight volume for the future prediction time period.

[0112] In some embodiments, the execution entity may input the target offline feature dataset and the target real-time feature dataset into a pre-trained real-time route freight volume generation model to generate a predicted route freight volume for the future prediction time period. The real-time route freight volume generation model is generated based on a model training method. The future prediction time period may be a pre-set time period. For example, the future prediction time period may be the day of the departure time and the two days thereafter.

[0113] Step 403 : performing volume correction on the predicted total volume of the route freight according to a pre-generated range of the total volume of the route freight, so as to generate a corrected total volume of the route freight.

[0114] In some embodiments, the execution entity may perform volume correction on the predicted total volume of the route freight according to a pre-generated total volume range of the route freight to generate a corrected total volume of the route freight. The total volume range of the route freight represents the predicted interval range of the total volume of the route freight.

[0115] For example, in response to determining that the predicted total freight volume of the route is less than the minimum value of the route total freight volume range, the corresponding value of the predicted total freight volume of the route is corrected to the minimum value of the route total freight volume range, thereby obtaining the corrected route total freight volume. In response to determining that the predicted total freight volume of the route is greater than the maximum value of the route total freight volume range, the corresponding value of the predicted total freight volume of the route is corrected to the maximum value of the route total freight volume range, thereby obtaining the corrected route total freight volume.

[0116] Optionally, the maximum value of the route freight volume range is the product of the upper bound of the predicted total volume and the first adjustment coefficient. Similarly, the minimum value of the route freight volume range is the product of the lower bound of the predicted total volume and the second adjustment coefficient. The upper bound and the lower bound of the predicted total volume can be the maximum and minimum total volumes in a historical time period before the current time (for example, the past month). The values ​​of the first adjustment coefficient and the second adjustment coefficient are set differently in different scenarios.

[0117] The specific coefficient value settings can be as follows:

[0118] (1) For big promotion scenarios, the route cargo volume will increase, the first adjustment coefficient can be set to 2, and the second adjustment coefficient can be set to 1.

[0119] (2) For daily scenarios, the route cargo volume is relatively stable, the first adjustment coefficient can be set to 1, and the second adjustment coefficient can be set to 1.

[0120] (3) For the Spring Festival scenario, the route cargo volume will decrease, the first adjustment coefficient can be set to 2, and the second adjustment coefficient can be set to 0.1.

[0121] In some optional implementations of some embodiments, after step 403, the steps further include:

[0122] The first step is to generate vehicle dispatch information for the target route based on the total freight volume of the corrected route, wherein the vehicle dispatch information can represent vehicle dispatch arrangement information for the target route.

[0123] As an example, the execution entity may generate vehicle dispatch information for the target route based on the total freight volume of the corrected route using a target dispatch table. The target dispatch table may represent the correspondence between the total freight volume of the corrected route and various vehicle dispatch arrangements.

[0124] The second step is to execute the dispatch processing of related vehicles according to the above vehicle dispatch information.

[0125] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: the route freight total volume generation method of some embodiments of the present disclosure can utilize a real-time route freight total volume generation model to accurately generate the corrected route freight total volume.

[0126] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a model training device. These device embodiments are similar to Figure 2 Corresponding to the method embodiments shown, the model training device can be specifically applied to various electronic devices.

[0127] like Figure 5 As shown, a model training device 500 includes: a first acquisition unit 501, a correction unit 502, a generation unit 503, and a training unit 504. The first acquisition unit 501 is configured to acquire an offline route freight dataset and a real-time route freight dataset; the correction unit 502 is configured to perform data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; the generation unit 503 is configured to generate an offline feature dataset and a real-time feature dataset based on the corrected offline route freight dataset and the corrected real-time route freight dataset; and the training unit 504 is configured to perform model training on an initial real-time route freight total volume generation model based on the offline feature dataset and the real-time feature dataset to generate a real-time route freight total volume generation model.

[0128] In some optional implementations of some embodiments, the offline route freight data in the above-mentioned offline route freight data set includes: historical actual shipment total volume and historical actual shipment vehicle information; and the correction unit 502 can be further configured to: for each offline route freight data in the above-mentioned offline route freight data set, perform the following first correction operation: generate the historical actual shipment vehicle total volume according to the historical actual shipment vehicle information corresponding to the above-mentioned offline route freight data; in response to determining that the historical actual shipment total volume corresponding to the above-mentioned offline route freight data is greater than the above-mentioned historical actual shipment vehicle total volume, correct the historical actual shipment total volume corresponding to the above-mentioned offline route freight data to the above-mentioned historical actual shipment vehicle total volume, and obtain the corrected offline route freight data.

[0129] In some optional implementations of some embodiments, the real-time route freight data in the above-mentioned real-time route freight data set includes: an estimated total volume of route freight for a target departure time; and the correction unit 502 can be further configured to: for each real-time route freight data in the above-mentioned real-time route freight data set, perform the following second correction operation: determine the corrected offline route freight data in the same period as the above-mentioned real-time route freight data as the target corrected offline route freight data; in response to the historical actual shipment total volume corresponding to the above-mentioned target corrected offline route freight data being smaller than the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data, correct the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data to the historical actual shipment total volume corresponding to the above-mentioned corrected offline route freight data.

[0130] In some optional implementations of some embodiments, the correction unit 502 can be further configured to: in response to determining that the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data is a negative number, correct the estimated total volume of route freight corresponding to the above-mentioned real-time route freight data to a target value.

[0131] In some optional implementations of some embodiments, the generation unit 503 can be further configured to: obtain an offline feature information set; extract an offline feature data set corresponding to the above-mentioned offline feature information set from the above-mentioned corrected offline route freight data set; for each corrected real-time route freight data in the above-mentioned corrected real-time route freight data set, perform the following generation steps: determine a subset of corrected real-time route freight data at the same time point as the above-mentioned corrected real-time route freight data; determine the quantile of the corresponding data size of the above-mentioned corrected real-time route freight data in the above-mentioned corrected real-time route freight data subset; determine the corrected offline route freight data in the above-mentioned corrected offline route freight data set whose corresponding data size quantile is the above-mentioned quantile as the quantile value; and determine the fusion information of the above-mentioned quantile and the above-mentioned quantile value as the real-time feature data for the above-mentioned corrected real-time route freight data.

[0132] In some optional implementations of some embodiments, the above-mentioned corrected real-time route freight data includes: the real-time route freight order quantity, and the above-mentioned quantile includes the first quantile; and the generation unit 503 can be further configured to: determine the quantile of the order quantity size of the real-time route freight order quantity corresponding to the above-mentioned corrected real-time route freight data in the quantile subset as the above-mentioned first quantile, wherein the above-mentioned quantile subset is the real-time route freight order subset corresponding to the above-mentioned corrected real-time route freight data subset.

[0133] In some optional implementations of some embodiments, the above-mentioned quantile value includes a first quantile value; and the generation unit 503 can be further configured to: determine the corrected offline route freight data set whose corresponding data size quantile in the above-mentioned corrected offline route freight data set is the above-mentioned first quantile point, as the first quantile value.

[0134] In some optional implementations of some embodiments, the above-mentioned corrected real-time route freight data includes: the estimated total volume of route freight, and the above-mentioned quantile includes the second quantile; and the generation unit 503 can be further configured to: determine the quantile of the volume size of the estimated total volume of route freight corresponding to the above-mentioned corrected real-time route freight data in the volume subset as the above-mentioned second quantile, wherein the above-mentioned volume subset is the estimated total volume subset of route freight corresponding to the above-mentioned corrected real-time route freight data subset.

[0135] In some optional implementations of some embodiments, the above-mentioned quantile value includes a second quantile value; and the generation unit 503 can be further configured to: determine the corrected offline route freight data set whose corresponding data size quantile in the above-mentioned corrected offline route freight data set is the above-mentioned second quantile point, as the second quantile value.

[0136] It is understandable that the units recorded in the model training device 500 and the reference Figure 2 The steps in the described method correspond to each other. Therefore, the operations, features and beneficial effects described above for the method are also applicable to the model training device 500 and the units contained therein, and will not be repeated here.

[0137] Further references Figure 6 As an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a route freight total volume generation device. These device embodiments are similar to Figure 4 Corresponding to the method embodiments shown, the route freight total volume generating device can be specifically applied to various electronic devices.

[0138] like Figure 6 As shown, a route freight volume generation device 600 includes: a second acquisition unit 601, a second input unit 602, and a correction unit 603. The second acquisition unit 601 is configured to acquire an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset; the second input unit 602 is configured to input the target offline feature dataset and the target real-time feature dataset into a pre-trained real-time route freight volume generation model to generate a predicted route freight volume for the future prediction time period, wherein the real-time route freight volume generation model is generated based on a model training method; and the correction unit 603 is configured to perform volume correction on the predicted route freight volume based on a pre-generated route freight volume range to generate a corrected route freight volume.

[0139] In some optional implementations of some embodiments, the route freight volume generating device 600 further includes an information generating unit and an execution unit (not shown). The information generating unit may be configured to generate vehicle scheduling information for the target route based on the corrected route freight volume. The execution unit may be configured to execute scheduling processing for the relevant vehicles based on the vehicle scheduling information.

[0140] It is understood that the units recorded in the route freight total volume generating device 600 are the same as those in the reference Figure 4Therefore, the operations, features and beneficial effects described above for the method are also applicable to the route freight total volume generating device 600 and the units included therein, and will not be repeated here.

[0141] Reference below Figure 7 , which shows an electronic device (eg, Figure 1 Schematic diagram of the structure of the electronic device 101)700. Figure 7 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0142] like Figure 7 As shown, the electronic device 700 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory 702 or a program loaded from a storage device 708 into a random access memory 703. Various programs and data required for the operation of the electronic device 700 are also stored in the random access memory 703. The processing device 701, the read-only memory 702, and the random access memory 703 are connected to each other via a bus 704. An input / output interface 705 is also connected to the bus 704.

[0143] Typically, the following devices may be connected to the input / output interface 705: an input device 706 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 707 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 708 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 709. The communication device 709 may allow the electronic device 700 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 7 The electronic device 700 is shown with various devices, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead. Figure 7 Each block shown in the figure may represent one device, or may represent multiple devices as needed.

[0144] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from a network via the communication device 709, or installed from the storage device 708, or installed from the read-only memory 702. When the computer program is executed by the processing device 701, the above-mentioned functions defined in the method of some embodiments of the present disclosure are performed.

[0145] It should be noted that in some embodiments of the present disclosure, the computer-readable medium mentioned above may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the 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 some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a 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. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0146] In some embodiments, the client and server can communicate using any currently known or future developed network protocol, such as HTTP (HyperText Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0147] The computer-readable medium may be included in the electronic device, or may exist independently and not be incorporated into the electronic device. The computer-readable medium carries one or more programs. When executed by the electronic device, the one or more programs cause the electronic device to: obtain an offline route freight dataset and a real-time route freight dataset; perform data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; generate an offline feature dataset and a real-time feature dataset based on the corrected offline route freight dataset and the corrected real-time route freight dataset; and perform model training on an initial real-time route freight total volume generation model based on the offline feature dataset and the real-time feature dataset to generate a real-time route freight total volume generation model. Obtain an offline feature dataset and a real-time feature dataset for the target route as the target offline feature dataset and the target real-time feature dataset; input the above target offline feature dataset and the above target real-time feature dataset into a pre-trained real-time route freight total volume generation model to generate a route freight forecast total volume for the above future prediction time period, wherein the above real-time route freight total volume generation model is generated based on the model training method corresponding to the first aspect; perform volume correction on the above route freight forecast total volume based on a pre-generated route freight total volume range interval to generate a corrected route freight total volume.

[0148] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0149] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of 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 than that 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 and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0150] The units described in some embodiments of the present disclosure may be implemented in software or hardware. The units described may also be provided in a processor. For example, they may be described as follows: a processor includes a first acquisition unit, a correction unit, a generation unit, and a training unit. The names of these units do not, in some cases, limit the units themselves. For example, the first acquisition unit may also be described as a "unit for acquiring an offline route freight dataset and a real-time route freight dataset."

[0151] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0152] Some embodiments of the present disclosure also provide a computer program product, including a computer program, which implements any of the above-mentioned model training methods or route freight total volume generation methods when executed by a processor.

[0153] The above description is only an illustration of some preferred embodiments of the present disclosure and the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A model training method, comprising: Obtain offline route freight dataset and real-time route freight dataset; performing data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; generating an offline feature dataset and a real-time feature dataset based on the revised offline route freight dataset and the revised real-time route freight dataset; The initial real-time route freight volume generation model is trained according to the offline feature data set and the real-time feature data set to generate a real-time route freight volume generation model.

2. The method according to claim 1, wherein The offline route freight data in the offline route freight data set includes: historical actual shipment total volume and historical actual shipment vehicle information; and The step of correcting the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset includes: For each offline route freight data in the offline route freight data set, the following first correction operation is performed: Generate the total volume of historical actual delivery vehicles based on the historical actual delivery vehicle information corresponding to the offline route freight data; In response to determining that the historical actual total shipment volume corresponding to the offline route freight data is greater than the historical actual total shipment vehicle volume, the historical actual total shipment volume corresponding to the offline route freight data is corrected to the historical actual total shipment vehicle volume to obtain corrected offline route freight data.

3. The method according to claim 2, wherein: The real-time route freight data in the real-time route freight data set includes: an estimated total volume of route freight for a target departure time; and The step of correcting the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset includes: For each real-time route freight data in the real-time route freight data set, the following second correction operation is performed: determining the revised offline route freight data in the same period as the real-time route freight data as the target revised offline route freight data; In response to the fact that the historical actual shipment total volume corresponding to the offline route freight data after the target correction is less than the route freight estimated total volume corresponding to the real-time route freight data, the route freight estimated total volume corresponding to the real-time route freight data is corrected to the historical actual shipment total volume corresponding to the corrected offline route freight data.

4. The method according to claim 3, wherein: After correcting the estimated total volume of route freight corresponding to the real-time route freight data to the historical actual total volume of shipment corresponding to the corrected offline route freight data in response to the target corrected total volume of historical actual shipments corresponding to the target corrected offline route freight data, the method further includes: In response to determining that the estimated total volume of route freight corresponding to the real-time route freight data is a negative number, the estimated total volume of route freight corresponding to the real-time route freight data is corrected to a target value.

5. The method according to claim 1, wherein The generating of an offline feature dataset and a real-time feature dataset based on the corrected offline route freight dataset and the corrected real-time route freight dataset includes: Obtain offline feature information set; extracting an offline feature data set corresponding to the offline feature information set from the corrected offline route freight data set; For each corrected real-time route freight data in the corrected real-time route freight data set, the following generating steps are performed: determining a subset of the revised real-time route freight data at a same time point as the revised real-time route freight data; Determine a quantile of the data size corresponding to the corrected real-time route freight data in the corrected real-time route freight data subset; Determine the corrected offline route freight data set having a corresponding data size quantile as the quantile, as a quantile value; The fusion information of the quantile point and the quantile value is determined as the real-time feature data for the corrected real-time route freight data.

6. The method according to claim 5, wherein: The corrected real-time route freight data includes: real-time route freight order quantity, the quantile includes a first quantile; and Determining the quantile of the data size corresponding to the corrected real-time route freight data in the corrected real-time route freight data subset includes: Determine a quantile of the order quantity of the real-time route freight order quantity corresponding to the corrected real-time route freight data in a sub-set of orders as the first quantile, wherein the sub-set of orders is a sub-set of the real-time route freight order quantity corresponding to the subset of the corrected real-time route freight data.

7. The method according to claim 6, wherein: The quantile values ​​include a first quantile value; and The determining of the corrected offline route freight data set having a corresponding data size quantile as the quantile as the corrected offline route freight data as the quantile value includes: The corrected offline route freight data set having a corresponding data size quantile as the first quantile is determined as the first quantile value.

8. The method according to claim 5, wherein The corrected real-time route freight data includes: an estimated total volume of route freight, the quantiles including a second quantile; and Determining the quantile of the data size corresponding to the corrected real-time route freight data in the corrected real-time route freight data subset includes: Determine the quantile of the volume of the estimated total volume of route freight corresponding to the corrected real-time route freight data in the volume subset as the second quantile, wherein the volume subset is the subset of the estimated total volume of route freight corresponding to the corrected real-time route freight data subset.

9. The method according to claim 8, wherein The quantile values ​​include a second quantile value; and The determining of the corrected offline route freight data set having a corresponding data size quantile as the quantile as the corrected offline route freight data as the quantile value includes: Determine the corrected offline route freight data set whose corresponding data size quantile is the second quantile as the second quantile value.

10. A method for generating a total freight volume of a route, comprising: Obtaining an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset; Inputting the target offline feature dataset and the target real-time feature dataset into a pre-trained real-time route freight volume generation model to generate a predicted route freight volume for the future prediction time period, wherein the real-time route freight volume generation model is generated based on the method of claims 1-9; Based on the pre-generated route freight total volume range, the predicted route freight total volume is volume-corrected to generate a corrected route freight total volume.

11. The method according to claim 10, wherein: The method further comprises: generating vehicle dispatch information for the target route based on the total freight volume of the corrected route; According to the vehicle dispatch information, dispatch processing of related vehicles is performed.

12. A model training device comprising: A first acquisition unit is configured to acquire an offline route freight dataset and a real-time route freight dataset; a correction unit configured to perform data correction on the offline route freight dataset and the real-time route freight dataset to generate a corrected offline route freight dataset and a corrected real-time route freight dataset; a generating unit configured to generate an offline feature dataset and a real-time feature dataset based on the corrected offline route freight dataset and the corrected real-time route freight dataset; The training unit is configured to perform model training on the initial real-time route freight volume generation model based on the offline feature data set and the real-time feature data set to generate a real-time route freight volume generation model.

13. A device for generating a total volume of freight transported on a route, comprising: a second acquiring unit configured to acquire an offline feature dataset and a real-time feature dataset for a target route as a target offline feature dataset and a target real-time feature dataset; a second input unit configured to input the target offline feature dataset and the target real-time feature dataset into a pre-trained real-time route freight total volume generation model to generate a predicted route freight total volume for the future prediction time period, wherein the real-time route freight total volume generation model is generated based on the method of claims 1 to 9; The correction unit is configured to perform volume correction on the predicted total volume of the route freight according to a pre-generated total volume range of the route freight to generate a corrected total volume of the route freight.

14. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11.

15. A computer-readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

16. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 11.

Images