Method, device, computer equipment and storage medium for predicting express delivery timeliness

Abstract

The present application relates to a method, device, computer equipment and storage medium for predicting the timeliness of express delivery. The present application can provide users with more accurate estimated arrival times of express delivery and predicted times of express delivery at various stages, which is conducive to reducing the number of inquiries and reminders from users and improving the service quality of express delivery companies. The present application can also calculate the quantity of express delivery distributed over time at different stages, improve the allocation of logistics resources, and thus reduce costs. The method includes determining the current state of the express delivery by obtaining online express delivery information; inputting the online express delivery information of the express delivery into a pre-built stage prediction model corresponding to the preset current state to obtain a stage prediction result; inputting the stage prediction result and the online express delivery information into a pre-built result credibility detection model to obtain a credibility detection result of the stage prediction result; obtaining the timeliness prediction result of the express delivery based on the stage prediction result and the credibility detection result; and sending the timeliness prediction result to the corresponding express delivery processing terminal.

Description

Technical Field

[0001] The present application relates to the field of logistics technology, and in particular to a method, device, computer equipment and storage medium for predicting the timeliness of express delivery. Background Art

[0002] With the development of e-commerce, logistics technology has continued to mature. Express parcels are not only required to be delivered accurately to the address required by the customer, but also to arrive as early as possible.

[0003] With existing logistics technology, some express delivery companies are unable to provide customers with accurate estimated arrival times during express delivery, resulting in customers being unable to effectively arrange pickup arrangements, which in turn affects users' pickup inspections. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment and storage medium for predicting the timeliness of express delivery in response to the above technical problems.

[0005] A method for predicting the timeliness of express delivery, the method comprising:

[0006] Obtain online express delivery information and determine the current status of the express delivery;

[0007] Inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result;

[0008] Inputting the stage prediction result and the online express delivery information into a pre-built result credibility detection model to obtain a credibility detection result of the stage prediction result;

[0009] Obtaining a timeliness prediction result of the express shipment according to the stage prediction result and the credibility detection result;

[0010] The timeliness prediction result is sent to the express delivery processing terminal corresponding to the express delivery.

[0011] In one embodiment, the pre-built stage prediction model includes a plurality of sub-models arranged in time sequence; the step of inputting the online express information of the express into the pre-built stage prediction model corresponding to the current state to obtain the stage prediction result includes:

[0012] Input the online express information into the sub-model of the time series where the current state is located to obtain the prediction result of the current sub-model;

[0013] Inputting the online express information and the prediction result of the current sub-model into the next sub-model of the sub-model in the time sequence of the current state, until the sub-model prediction result of the last sub-model of the multiple sub-models is obtained;

[0014] The prediction result of the stage is obtained according to the sub-model prediction results of each sub-model.

[0015] In one embodiment, the stage prediction model includes a receipt sub-model, a transit sub-model, and a delivery sub-model arranged in time sequence; before inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain the stage prediction result, the method further includes:

[0016] The express life cycle is divided into the collection stage, the transit stage and the delivery stage according to the time label;

[0017] Obtain attribute information samples and logistics operation information samples of multiple express delivery samples;

[0018] Acquire training data from the attribute information samples and the logistics operation information samples by feature selection;

[0019] Filtering the training data according to the time tags to obtain the training data of each stage matching the receiving stage, the transfer stage and the delivery stage;

[0020] The training data of each stage are used to train the receiving sub-model, the intermediate sub-model and the dispatching sub-model to be trained respectively, so as to construct the receiving sub-model, the intermediate sub-model and the dispatching sub-model.

[0021] In one embodiment, before inputting the stage prediction result and the online express information into a pre-built credibility model to obtain a credibility result corresponding to the current state, the method further includes:

[0022] Using the receipt sub-model, the intermediate sub-model and the delivery sub-model, obtaining stage prediction result samples of the plurality of express delivery samples;

[0023] Obtaining real stage samples of the multiple express samples;

[0024] Obtaining a stage prediction error according to the stage prediction result sample and the true stage sample;

[0025] Encoding the attribute information samples of the plurality of express delivery samples according to the stage prediction error and a preset error threshold to obtain credibility training data;

[0026] The credibility detection model to be trained is trained using the credibility training data to construct the credibility detection model.

[0027] In one embodiment, determining the current status of the express shipment includes:

[0028] Obtaining the delivery characteristic information of the express delivery;

[0029] Determining whether the express shipment is in the receipt stage according to the receipt characteristic information;

[0030] If so, it is determined that the current status of the express is a receiving status.

[0031] In one embodiment, after determining whether the express delivery is in the delivery stage according to the delivery feature information, the method further includes:

[0032] If not, then obtain the current outlet information where the express is located;

[0033] Determine whether the current network point information is an originating network point or a transit network point;

[0034] If so, it is determined that the current status of the express is in transit.

[0035] In one embodiment, after obtaining the current network point information of the express delivery, the method further includes:

[0036] If it is found that the current network point information is consistent with the express destination, it is determined that the current status of the express is the delivery status.

[0037] A device for predicting the timeliness of express delivery, comprising:

[0038] The express information acquisition module is used to obtain online express information and determine the current status of the express;

[0039] A result prediction module, used for inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result;

[0040] A credibility detection module, used for inputting the stage prediction result and the online express information into a pre-built result credibility detection model to obtain a credibility detection result of the stage prediction result;

[0041] A timeliness prediction result determination module is used to obtain the timeliness prediction result of the express shipment according to the stage prediction result and the credibility detection result;

[0042] The timeliness prediction result sending module is used to send the timeliness prediction result to the express delivery processing terminal corresponding to the express delivery.

[0043] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above-mentioned express delivery time prediction method embodiment when executing the computer program.

[0044] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the above-mentioned express delivery time prediction method embodiment.

[0045] The above-mentioned express delivery timeliness prediction method, device, computer equipment and storage medium determine the current state of the express delivery by obtaining online express delivery information; input the online express delivery information of the express delivery into a pre-built stage prediction model corresponding to the preset current state to obtain the stage prediction result; input the stage prediction result and the online express delivery information into a pre-built result credibility detection model to obtain the credibility detection result of the above-mentioned stage prediction result; obtain the timeliness prediction result of the express delivery according to the stage prediction result and the credibility detection result; send the timeliness prediction result to the express delivery processing end corresponding to the above-mentioned express delivery. The method can provide users with more accurate estimated arrival time of express delivery and predicted time of express delivery at each stage, which is conducive to reducing the number of user inquiries and reminders, and improving the service quality of express delivery companies. Furthermore, the method can also be applied to calculate the quantity of express delivery distributed over time at different stages according to the predicted arrival time of express delivery at different stages, so as to arrange relevant resources, improve the allocation of logistics resources, thereby reducing costs, speeding up express delivery transportation, and ensuring that express delivery arrives as soon as possible. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is an application environment diagram of a method for predicting the timeliness of express delivery in one embodiment;

[0047] Figure 2 A schematic diagram of a flow chart of a method for predicting the timeliness of express delivery in one embodiment;

[0048] Figure 3 A schematic diagram of a prediction process of a stage prediction model in an embodiment;

[0049] Figure 4 A schematic diagram of a training process of a stage prediction model in an embodiment;

[0050] Figure 5 A schematic diagram of a training process of a result credibility detection model in one embodiment;

[0051] Figure 6 A schematic diagram of a flow chart of determining the status of a parcel in one embodiment;

[0052] Figure 7 A structural block diagram of a device for predicting the timeliness of express delivery in one embodiment;

[0053] Figure 8 is an internal structure diagram of a computer device in one embodiment;

[0054] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in another embodiment. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0056] The express delivery time prediction method provided in this application can be applied to Figure 1 In the application environment shown in the figure, the terminal 101 communicates with the server 102 through the network. The terminal 101 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, and portable wearable devices, and the server 102 can be implemented by an independent server or a server cluster composed of multiple servers. The server 102 can be a server located in each network point, used to collect all express information of the current network point.

[0057] In one embodiment, Figure 2 As shown in the figure, a method for predicting the timeliness of express delivery is provided. Figure 1 Taking the server 102 in the example as an example, the following steps are included:

[0058] Step S201, obtaining online express delivery information and determining the current status of the express delivery.

[0059] Among them, online express information is only real-time information of express, including the attribute information and logistics operation information of express. The attribute information includes the type of goods, weight, mailing address, receiving address, time type (such as expedited, ordinary), etc. Logistics operation information refers to the information processed by the express company, such as the current network type, dispatch time, arrival time, collection time, etc. The current status refers to the stage status divided in advance according to the life cycle of the express, such as delivery status, transit status and delivery status.

[0060] Specifically, the server 102 obtains the property information and operation information of the current express in real time, first parses the property information and operation information, then aggregates the property information and operation information of the express, and makes a status judgment based on the aggregated data to obtain the current status of the express.

[0061] Step S202, inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result.

[0062] Among them, the pre-built stage prediction model is used to predict the estimated arrival time of express delivery at various stages. For example, in the receiving stage, the dispatch time of the express delivery can be predicted. In the transit stage, the estimated time for the express delivery to arrive at the delivery point can be predicted. In the delivery stage, the dispatch time and the estimated final arrival time of the express delivery can be predicted.

[0063] Specifically, the online express information of the express is input into the pre-built stage prediction model to obtain the stage prediction result. For example, if the express is in the receiving state, the data is sorted, the current state is created, the prediction corresponding to the current state is triggered, and the online express information is input into the receiving model to predict the receiving stage time; if the express is in the transit state, the prediction corresponding to the current state is triggered, and the online express information is input into the transit model to predict the estimated arrival time of the express at the delivery point in the transit stage; if the express is in the delivery state, the prediction corresponding to the current state is triggered, and the online express information is input into the delivery model to predict the delivery stage time.

[0064] Step S203, inputting the above-mentioned stage prediction results and the above-mentioned online express delivery information into a pre-built result credibility detection model to obtain the credibility detection results of the above-mentioned stage prediction results.

[0065] The definition of credibility is an estimation of whether the prediction result is accurate. For example, if the credibility is 60%, it indicates that there is a 60% probability that the prediction result is accurate.

[0066] Specifically, the above-mentioned stage prediction results and the above-mentioned online express delivery information are integrated, and input into the result credibility detection model to obtain the credibility of the stage prediction results of each stage.

[0067] Step S204, obtaining the timeliness prediction result of the above express delivery according to the stage prediction result and the credibility detection result.

[0068] Among them, the timeliness prediction results include the stage prediction results of each stage and the corresponding credibility.

[0069] Specifically, the results output by the prediction model in the integration phase and the results output by the result credibility detection model are combined to serve as the timeliness prediction result of the express delivery.

[0070] Step S205, sending the timeliness prediction result to the express delivery processing terminal corresponding to the above express delivery.

[0071] Specifically, the above-mentioned timeliness prediction result is sent to the terminal 101, and the terminal 101 can be a terminal used by a courier or a terminal used by a user.

[0072] In the above embodiment, the current state of the express is determined by obtaining online express information; the online express information of the express is input into a pre-built stage prediction model corresponding to the preset current state to obtain a stage prediction result; the stage prediction result and the online express information are input into a pre-built result credibility detection model to obtain a credibility detection result of the above stage prediction result; the timeliness prediction result of the express is obtained according to the stage prediction result and the credibility detection result; the timeliness prediction result is sent to the express processing terminal corresponding to the above express. This method can provide users with more accurate estimated arrival time of the express and the predicted time of the express at each stage, and provide the credibility of the prediction result, which is conducive to reducing the number of inquiries and reminders from users, and at the same time, it also helps to improve the service quality of the express company.

[0073] In one embodiment, if Figure 3 As shown, Figure 3 The schematic diagram of the prediction process of the stage prediction model is shown. The pre-built stage prediction model includes multiple sub-models arranged in time sequence. The step S202 includes:

[0074] Input the online express information into the sub-model of the time series where the current state is located to obtain the prediction result of the current sub-model; input the online express information and the prediction result of the current sub-model into the next sub-model of the sub-model of the time series where the current state is located, until the sub-model prediction result of the last sub-model of multiple sub-models is obtained; obtain the stage prediction result according to the sub-model prediction results of each sub-model.

[0075] Among them, the pre-built stage prediction model includes multiple sub-models arranged in time sequence. For example, it can be divided into a receiving sub-model, a transit sub-model and a delivery sub-model according to the express life cycle.

[0076] Specifically, if the express is in the state of starting to receive, sort out the data, create a state type, trigger the prediction, input the receiving sub-model to predict the receiving stage time, then transform the predicted receiving time, integrate the test data, input the intermediate sub-model to predict the transit stage time, then transform the transit prediction time, integrate the test data, input the dispatch sub-model to predict the dispatch stage time, process the dispatch prediction time to obtain the express delivery time, and finally realize the full-process time efficiency prediction of the express. If the express is in the state of starting to transfer, sort out the data, create a state type or update the state type, trigger the prediction, input the intermediate sub-model to predict the transit stage time, then transform the transit prediction time, integrate the test data, input the dispatch sub-model to predict the dispatch stage time, and process the dispatch prediction time to obtain the express delivery time. If the express is in the state of arriving at the delivery point, sort out the data, create a state type or update the state type, trigger the prediction, input the dispatch sub-model to predict the dispatch stage time, and process the dispatch prediction time to obtain the express delivery time.

[0077] In addition, if the express is not in any of these three states, no processing is performed and the express information of the next express is processed.

[0078] In the above embodiment, according to the current status of the express, the express information is input into the sub-model corresponding to the current status, and then into the next sub-model in turn, to obtain the stage prediction results corresponding to each model. Compared with the traditional method of only generally predicting the final arrival time of the express, it is more refined, and the prediction information is updated at any time as the status of the express changes, providing users and couriers with more accurate time reference information.

[0079] In one embodiment, if Figure 4 As shown, Figure 4 The training process diagram of the stage prediction model is shown. The stage prediction model includes a receiving sub-model, a transit sub-model and a dispatching sub-model arranged in time sequence. Before the above step S202, the method further includes:

[0080] Step S401, dividing the express life cycle into a receiving stage, a transit stage and a delivery stage according to the time tag.

[0081] Specifically, the entire life cycle of express delivery starts from when it is collected by the sender and ends when it is delivered to the recipient. According to some landmark events in the process, the process is divided into three stages: the collection stage, the transit stage and the delivery stage.

[0082] Step S402, obtaining attribute information samples and logistics operation information samples of multiple express delivery samples;

[0083] Among them, express delivery samples refer to historical express delivery, such as express delivery that has completed its entire life cycle. Attribute information samples include the type of goods, weight, shipping address, receiving address, timeliness type (such as expedited, ordinary), etc. of historical express delivery. Logistics operation information samples refer to information processed by the express delivery company, such as the current network type, dispatch time, arrival time, and collection time.

[0084] Specifically, the mobile phone contains attribute information samples and logistics operation information samples of historical express parcels for approximately two weeks.

[0085] Step S403, obtaining training data from attribute information samples and logistics operation information samples through feature selection.

[0086] Specifically, firstly, the express data with abnormalities in the entire life cycle, such as detention, transportation abnormality, incomplete life cycle, etc., are cleared, and then the cleaned data is matched with the express attribute information, and the data is grouped according to the division of the above three stages, and the data of each stage are transformed by time information (such as: life cycle duration, duration of each stage, whether it is a working day, start hour of each stage, etc.), location information transformation (such as: location type, mailing distance, origin city, destination city, etc.), time information transformation and other transformations; finally, the random forest model is used to select the top four features with high feature importance; the remaining features are verified by the correlation test method, and the features with relevant and relatively concentrated value ranges (such as weight) are selected; the features with weak correlation and relatively dispersed value ranges (such as delivery outlets, receiving outlets, etc.) are clustered for this type of features, and the influence of different feature values ​​on the duration is analyzed, and the feature values ​​with obvious influence are added to the features; then the data are grouped according to the origin city-destination city, and the data with abnormal duration in the group are removed; finally, the continuous features are discretized, and the category features are encoded to obtain the final training data.

[0087] Step S404, filtering the training data according to the time tags to obtain the training data of each stage matching the receiving stage, the terminal stage and the delivery stage.

[0088] Specifically, the training data of each stage is obtained by filtering the training data according to the time node information.

[0089] Step S405, respectively train the to-be-trained item receiving sub-model, the intermediate sub-model and the item dispatching sub-model using the training data of each stage, and construct the above item receiving sub-model, the intermediate sub-model and the item dispatching sub-model.

[0090] Specifically, the offline receiving sub-model, offline intermediate rotor model, and offline delivery sub-model based on linear regression are trained respectively with the duration of each stage as the label. The prediction effect of the model is verified using multiple statistical indicators such as the mean absolute error, the percentile of absolute error, and the proportion of items under different absolute error thresholds, and finally the trained sub-models are obtained.

[0091] In the above embodiment, historical express delivery sample data is collected to train sub-models at different stages to obtain trained sub-models, which is beneficial for subsequent use of these trained sub-models to perform predictions at various stages on real-time express delivery data.

[0092] In one embodiment, if Figure 5 As shown, Figure 5 The training process diagram of the result credibility detection model is shown. Before the above step S203, it also includes:

[0093] The stage prediction result samples of multiple express delivery samples are obtained by using the receiving sub-model, the intermediate sub-model and the dispatching sub-model; the real stage samples of the multiple express delivery samples are obtained; the stage prediction error is obtained according to the stage prediction result samples and the real stage samples; the attribute information samples of the multiple express delivery samples are encoded according to the stage prediction error and the preset error threshold to obtain the credibility training data; the credibility training data is used to train the credibility detection model to be trained, and the credibility detection model is constructed.

[0094] Among them, the stage prediction result samples refer to the stage prediction result samples obtained by predicting the historical express samples through the above-mentioned receiving sub-model, intermediate sub-model and dispatching sub-model; the real stage samples refer to the real time nodes corresponding to the historical express samples in the actual delivery process.

[0095] Specifically, historical express samples are obtained, and the receiving sub-model, the intermediate sub-model, and the dispatching sub-model are used to predict the historical express samples respectively, and the stage prediction result samples of each stage are obtained; the real stage samples of the historical express samples are obtained from the database, the prediction error between the stage pre-stored result samples and the real stage samples is calculated, and the error threshold is set. According to the error threshold, the prediction error is converted into 0 and 1 labels, where 1 indicates that the prediction is accurate below the threshold, and 0 indicates that the prediction is inaccurate below the threshold. Other information refers to the feature selection processing method in the stage prediction model, selects features, performs time information transformation, and class data encoding to obtain training data. Since the output of the credibility is a value in the range of 0-1, the logistic regression model is used as the result credibility detection model for training. Due to the huge amount of sample data (tens of millions), it is impossible to fully verify the test results. Therefore, a small range verification is adopted. The test samples within a certain probability range are taken, and the mean of the test samples in the range is calculated to see whether the sample mean is within the probability range. If it is within the range, and the error between the model output mean and the sample mean is less than the threshold, it means that the model predicts accurately within the probability range. Taking multiple intervals, the model effect can be approximately estimated.

[0096] In the above embodiment, the error between the predicted results and the actual results of historical express shipments at various stages is used to train the result credibility detection model, thereby obtaining a trained result credibility detection model, which lays a data foundation for the subsequent credibility detection of the stage prediction results of the express shipments.

[0097] In one embodiment, if Figure 6 As shown, Figure 6 The flowchart of determining the status of a shipment is shown. The determination of the current status of the shipment in step S201 includes:

[0098] Acquire the receiving characteristic information of the express; determine whether the express is in the receiving stage according to the receiving characteristic information; if so, determine that the current state of the express is the receiving state.

[0099] Among them, the receiving characteristic information includes the receiving code.

[0100] Specifically, the express delivery information and logistics operation information of the express delivery are obtained, wherein the express delivery information includes the waybill origin, destination outlet, distribution transfer yard, and routing information; the logistics operation information includes the waybill number, current operation time, current outlet, vehicle destination and other information. After the express delivery information and logistics operation information are aggregated and processed, they are stored in the database and bound to the receipt code for storage. After scanning the receipt code on the express delivery with the bar gun, it is determined whether the receipt code is consistent with the preset information of the receipt stage, for example, whether the receipt code contains 50 / 54 / 51. If so, the receipt sub-model prediction is triggered, and the receipt sub-model is used to determine the stage prediction result of the express delivery at the receipt stage based on the online express delivery information.

[0101] In the above embodiment, the prediction of the receiving sub-model is triggered by the bar gun code according to the preset code, which is conducive to automatically obtaining the prediction result of the express delivery.

[0102] In one embodiment, after determining whether the express delivery is in the delivery stage according to the delivery feature information, the method further includes:

[0103] If not, obtain the current network point information of the express delivery; determine whether the current network point information is an originating network point or a transit network point; if so, determine that the current status of the express delivery is a transit status.

[0104] Specifically, if it is judged that the receiving code is inconsistent with the preset information of the preset receiving stage, for example, if it is judged that the current outlet of the express is the originating outlet, and the receiving code is 30 / 302 (loading / departing), and there is no transit status label, then the transit sub-model prediction is triggered. In addition, if the current outlet of the express is a transit outlet and the transit status label does not exist, the transit sub-model prediction may also be triggered.

[0105] In the above embodiment, whether the express is in the transit stage is determined by the current outlet of the express, which is conducive to triggering the transit sub-model prediction.

[0106] In one embodiment, after obtaining the current branch information of the express, the method further includes: if it is found that the current branch information is consistent with the destination of the express, determining that the current status of the express is a delivery status.

[0107] Specifically, after obtaining the current network point information where the express is located, if the current network point information is consistent with the destination of the express (such as the vehicle destination), it is determined that the current status of the express is the delivery status.

[0108] The above embodiment triggers the prediction of the delivery sub-model by judging whether the current outlet of the express is consistent with the destination of the express, which is helpful to clearly distinguish the prediction results of each stage; at the same time, a delivery status label can be established to avoid repeated triggering.

[0109] It should be understood that although Figure 1-6 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-6 At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0110] In one embodiment, Figure 7 As shown, a device 700 for predicting the timeliness of express delivery is provided, comprising: an express delivery information acquisition module 701, a result prediction module 702, a credibility detection module 703, a timeliness prediction result determination module 704 and a timeliness prediction result sending module 705, wherein:

[0111] The express information acquisition module 701 is used to acquire online express information and determine the current status of the express;

[0112] The result prediction module 702 is used to input the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result;

[0113] Credibility detection module 703, used for inputting the stage prediction result and the online express information into a pre-built result credibility detection model to obtain a credibility detection result of the stage prediction result;

[0114] The timeliness prediction result determination module 704 is used to obtain the timeliness prediction result of the express shipment according to the stage prediction result and the credibility detection result;

[0115] The timeliness prediction result sending module 705 is used to send the timeliness prediction result to the express delivery processing terminal corresponding to the express delivery.

[0116] In one embodiment, the above-mentioned result prediction module 702 is also used to: input the online express delivery information into the sub-model of the time series where the current state is located to obtain the prediction result of the current sub-model; input the online express delivery information and the prediction result of the current sub-model into the next sub-model of the sub-model of the time series where the current state is located, until the sub-model prediction result of the last sub-model of the multiple sub-models is obtained; and obtain the stage prediction result according to the sub-model prediction results of each sub-model.

[0117] In one embodiment, it also includes a stage prediction model training unit, which is used to: divide the express life cycle into a receiving stage, a transit stage and a delivery stage according to time tags; obtain attribute information samples and logistics operation information samples of multiple express samples; obtain training data from the attribute information samples and logistics operation information samples through feature selection; filter the training data according to the time tags to obtain training data for each stage that matches the receiving stage, the transit stage and the delivery stage; use the training data for each stage to train the receiving sub-model, the transit sub-model and the delivery sub-model to be trained, respectively, to construct the receiving sub-model, the transit sub-model and the delivery sub-model.

[0118] In one embodiment, it also includes a result credibility detection model training unit, which is used to: use the receiving sub-model, the intermediate sub-model and the dispatching sub-model to obtain the stage prediction result samples of the multiple express samples; obtain the real stage samples of the multiple express samples; obtain the stage prediction error according to the stage prediction result samples and the real stage samples; encode the attribute information samples of the multiple express samples according to the stage prediction error and a preset error threshold to obtain credibility training data; use the credibility training data to train the credibility detection model to be trained, and construct the credibility detection model.

[0119] In one embodiment, the express information acquisition module 701 is further used to: acquire the receiving characteristic information of the express; determine whether the express is in the receiving stage according to the receiving characteristic information; if so, determine that the current state of the express is the receiving state.

[0120] In one embodiment, the express information acquisition module 701 is further used to: if not, obtain the current branch information of the express; determine whether the current branch information is the originating branch or the transit branch; if so, determine that the current state of the express is the transit state.

[0121] In one embodiment, the express information acquisition module 701 is further used to: if it is found that the current network point information is consistent with the express destination, determine that the current status of the express is the delivery status.

[0122] For the specific definition of the express delivery time prediction device, please refer to the definition of the express delivery time prediction method above, which will not be repeated here. Each module in the above-mentioned express delivery time prediction device can be implemented in whole or in part by software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0123] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store online express information and historical express information. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for predicting the timeliness of express delivery is implemented.

[0124] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for predicting the timeliness of express delivery is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0125] Those skilled in the art will understand that Figure 8-9The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0126] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps in the express delivery time prediction method as described above are implemented.

[0127] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the express delivery time prediction method as described above are implemented.

[0128] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0129] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0130] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for predicting the timeliness of express delivery, characterized in that: The method comprises: Obtaining online express information and determining the current status of the express; the current status is a stage status pre-divided according to the life cycle of the express; the current status includes receiving status, transit status and delivery status; Inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result; the pre-built stage prediction model is used to predict the estimated arrival time of the express at each stage; the stages include the collection stage, the transit stage and the delivery stage; The stage prediction result and the online express information are input into a pre-constructed result credibility detection model to obtain a credibility detection result of the stage prediction result; the credibility is an estimation of whether the prediction result is accurate; the result credibility detection model is constructed by training the credibility detection model to be trained using credibility training data; the credibility training data is obtained by encoding the attribute information samples of the stage prediction error and the preset error threshold for multiple express samples; Obtaining a timeliness prediction result of the express shipment according to the stage prediction result and the credibility detection result; The timeliness prediction result is sent to the express delivery processing terminal corresponding to the express delivery.

2. The method according to claim 1, characterized in that The pre-built stage prediction model includes a plurality of sub-models arranged in time sequence; the step of inputting the online express information of the express into the pre-built stage prediction model corresponding to the current state to obtain a stage prediction result includes: Input the online express information into the sub-model of the time series where the current state is located to obtain the prediction result of the current sub-model; Inputting the online express information and the prediction result of the current sub-model into the next sub-model of the sub-model in the time sequence of the current state, until the sub-model prediction result of the last sub-model of the multiple sub-models is obtained; The prediction result of the stage is obtained according to the sub-model prediction results of each sub-model.

3. The method according to claim 2, characterized in that The stage prediction model includes a receiving sub-model, a transit sub-model and a dispatching sub-model arranged in time sequence; before inputting the online express information of the express into the pre-built stage prediction model corresponding to the current state to obtain the stage prediction result, the method further includes: The express life cycle is divided into the collection stage, the transit stage and the delivery stage according to the time label; Obtain attribute information samples and logistics operation information samples of multiple express delivery samples; Acquire training data from the attribute information samples and the logistics operation information samples by feature selection; Filtering the training data according to the time tags to obtain the training data of each stage matching the receiving stage, the transfer stage and the delivery stage; The training data of each stage are used to train the receiving sub-model, the intermediate sub-model and the dispatching sub-model to be trained respectively, so as to construct the receiving sub-model, the intermediate sub-model and the dispatching sub-model.

4. The method according to claim 3, characterized in that Before inputting the stage prediction result and the online express delivery information into a pre-built result credibility detection model to obtain the credibility detection result of the stage prediction result, the method further includes: Using the receipt sub-model, the intermediate sub-model and the delivery sub-model, obtaining stage prediction result samples of the plurality of express delivery samples; Obtaining real stage samples of the multiple express delivery samples; Obtaining a stage prediction error according to the stage prediction result sample and the true stage sample; Encoding the attribute information samples of the plurality of express delivery samples according to the stage prediction error and a preset error threshold to obtain credibility training data; The credibility detection model to be trained is trained using the credibility training data to construct the credibility detection model.

5. The method according to claim 1, characterized in that The determining the current status of the express shipment includes: Obtaining the delivery characteristic information of the express delivery; Determining whether the express shipment is in the receipt stage according to the receipt characteristic information; If so, it is determined that the current status of the express is a receiving status.

6. The method according to claim 5, characterized in that After determining whether the express is in the receiving stage according to the receiving feature information, the method further includes: If not, then obtain the current branch information where the express is located; Determine whether the current network point information is an originating network point or a transit network point; If so, it is determined that the current state of the express is in transit.

7. The method according to claim 6, characterized in that After obtaining the current network point information of the express delivery, the method further includes: If it is found that the current network point information is consistent with the express destination, it is determined that the current status of the express is the delivery status.

8. A device for predicting the timeliness of express delivery, characterized in that: The device comprises: The express information acquisition module is used to obtain online express information and determine the current state of the express; the current state is a stage state divided in advance according to the life cycle of the express; the current state includes the receiving state, the transit state and the delivery state; A result prediction module, used for inputting the online express information of the express into a pre-built stage prediction model corresponding to the current state to obtain a stage prediction result; the pre-built stage prediction model is used for predicting the estimated arrival time of the express at each stage; the stages include the collection stage, the transit stage and the delivery stage; A credibility detection module is used to input the stage prediction result and the online express information into a pre-constructed result credibility detection model to obtain a credibility detection result of the stage prediction result; the credibility is an estimation of whether the prediction result is accurate; the result credibility detection model is constructed by training the credibility detection model to be trained using credibility training data; the credibility training data is obtained by encoding the attribute information samples of the stage prediction error and the preset error threshold for multiple express samples; A timeliness prediction result determination module is used to obtain the timeliness prediction result of the express shipment according to the stage prediction result and the credibility detection result; The timeliness prediction result sending module is used to send the timeliness prediction result to the express delivery processing terminal corresponding to the express delivery.

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

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

Images