Navigation distance estimation method and device, storage medium and equipment

The navigation distance estimation method uses historical ride order data and linear regression with city and time features to enhance accuracy, addressing the inaccuracy of existing methods and improving customer satisfaction.

CN119090587AInactive Publication Date: 2024-12-06BEIJING XINGYUN ONLINE SOFTWARE DEVELOPMENT CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411183601.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-06
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing methods for estimating navigation distance in online ride-hailing services are inaccurate due to urban road structures and lack of consideration for city characteristics, leading to reduced accuracy and stability.

Method used

A navigation distance estimation method using historical ride order data, including data preprocessing, linear regression modeling with city and time features, and real-time performance monitoring to adjust for different city conditions.

Benefits of technology

Improves navigation distance estimation accuracy by 34% compared to existing methods, enhancing the reputation and customer satisfaction of online ride-hailing platforms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090587A_ABST
    Figure CN119090587A_ABST
Patent Text Reader

Abstract

The invention provides a navigation distance estimation method and device, a storage medium and equipment. The method comprises the following steps: acquiring order information of a plurality of historical riding orders; performing data processing on the order information to obtain processed effective order data; constructing a navigation distance estimation model, and training the navigation distance estimation model by using the effective order data to obtain a trained target navigation distance estimation model; and estimating the navigation distance of the current riding order by using the target navigation distance estimation model. According to the invention, the linear regression model is used and city characteristics are considered, and adjustment is carried out according to the characteristics of different cities, so that actual road and traffic conditions are reflected more accurately, more accurate distance estimation is realized, and an actual and ubiquitous problem in the online car-hailing industry is solved. The reputation and customer satisfaction of the online car-hailing platform are greatly improved, and the method has very high practical value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle navigation, and in particular to a navigation distance estimation method and device, a storage medium and equipment. Background Art

[0002] A key problem in the online ride-hailing industry is how to accurately estimate the navigation distance between a passenger's departure and destination. The solution to this problem directly affects the accuracy of passenger fares, which in turn affects the reputation and customer satisfaction of the online ride-hailing platform. However, in some cases, due to various reasons (such as network problems, map interface limitations, etc.), it is impossible to obtain the actual navigation distance.

[0003] At present, the following methods are used to estimate the navigation distance in the online car-hailing industry:

[0004] (1) Distance measurement method based on Euclidean distance. In order to improve the accuracy, the calculation method of "Euclidean distance × 1.4" is usually used. However, this method has some obvious disadvantages: Inadaptability to urban roads: Euclidean distance calculates the straight-line distance between two points, which may not be accurate in the grid-like or irregular road structure of the city; Lack of robustness: Euclidean distance is more sensitive to outliers or noise, which may affect the accuracy and stability of the model.

[0005] (2) Estimation method based on Manhattan distance. Manhattan distance is a commonly used distance estimation method, which calculates the sum of the distances between two points in various dimensions. However, this method has several limitations: it does not consider actual road and traffic conditions; it cannot adapt to the characteristics of different cities or regions.

[0006] (3) Calculation methods that simply use linear regression algorithms without considering city characteristics. Although linear regression is a powerful and widely used statistical method, in the navigation distance estimation of the online car-hailing industry, if linear regression is simply used without considering city characteristics, it may face the following problems: insufficient generalization ability in different cities, poor interpretability due to different performance capabilities; different city characteristics lead to reduced accuracy and limited practicality; insufficient model data utility and insufficient complexity. Summary of the invention

[0007] To this end, the present invention provides a navigation distance estimation method and device, a storage medium, and a device, aiming to solve the technical problem in the prior art that the distance between two points cannot be accurately estimated when it is difficult to obtain the actual navigation distance.

[0008] To achieve the above objectives, the present invention adopts the following technical solutions:

[0009] According to a first aspect of the present invention, the present invention provides a navigation distance estimation method, characterized in that the method comprises:

[0010] Get order information of multiple historical ride orders;

[0011] Performing data processing on the order information to obtain processed valid order data;

[0012] Constructing a navigation distance estimation model, and training the navigation distance estimation model using the valid order data to obtain a trained target navigation distance estimation model;

[0013] The target navigation distance estimation model is used to estimate the navigation distance of the current ride order.

[0014] Optionally, the order information includes at least one of city information, order time, departure coordinates, destination coordinates and actual navigation distance.

[0015] Optionally, the performing data processing on the order information to obtain processed valid order data includes:

[0016] Performing data preprocessing on the order information to obtain valid order information; the data preprocessing includes at least one of removing duplicate records, clearing missing values, and processing outliers;

[0017] Grouping the valid order information according to the city information to obtain multiple pieces of valid order information corresponding to each target city;

[0018] Calculate the Euclidean distance between the coordinates of the departure place and the coordinates of the destination in each valid order information, and add the Euclidean distance to the valid order information;

[0019] The valid order information is normalized and / or standardized to obtain valid order data.

[0020] Optionally, the constructing of the navigation distance estimation model includes:

[0021] Distance features and time features are selected as model features to build a navigation distance estimation model based on linear regression;

[0022] The distance feature includes Euclidean distance and / or actual navigation distance; the time feature includes hour information, week information and / or weekend information.

[0023] Optionally, the navigation distance estimation model based on linear regression is expressed using the following formula:

[0024] y=β0+β1x+∈

[0025] Among them, y represents the actual navigation distance; x represents the Euclidean distance; β0 and β1 are model parameters respectively; ∈ is the error term.

[0026] Optionally, the using the valid order data to train the navigation distance estimation model to obtain a trained target navigation distance estimation model includes:

[0027] For each of the city information, the corresponding valid order data is divided into a training sample set and a test sample set;

[0028] Using the training sample set to train the navigation distance estimation model;

[0029] The trained navigation distance estimation model is tested using the test sample set, and the navigation distance estimation model is cross-validated by calculating the model performance index to obtain a navigation distance estimation model that meets the performance standard as the target navigation distance estimation model corresponding to each of the city information that has been trained;

[0030] The model performance index includes at least one of mean square error, determination coefficient and mean absolute error.

[0031] Optionally, the method further comprises:

[0032] The target navigation distance estimation model is subjected to real-time performance monitoring, and abnormal warning and / or version updating are performed on the target navigation distance estimation model based on the model performance.

[0033] According to a second aspect of the present invention, the present invention provides a navigation distance estimation device, the device comprising:

[0034] An information acquisition module, used to acquire order information of multiple historical ride orders;

[0035] A data processing module, used to process the order information to obtain processed valid order data;

[0036] A model building module, used to build a navigation distance estimation model, and train the navigation distance estimation model using the valid order data to obtain a trained target navigation distance estimation model;

[0037] The distance estimation module is used to estimate the navigation distance of future ride orders using the target navigation distance estimation model.

[0038] According to a third aspect of the present invention, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the navigation distance estimation method as described in any one of the first aspect of the present invention.

[0039] According to a fourth aspect of the present invention, the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements a navigation distance estimation method as described in any one of the first aspects of the present invention when executing the computer program.

[0040] The present invention adopts the above technical solution and has at least the following beneficial effects:

[0041] Through the scheme of the present invention, the order information of multiple historical ride orders is obtained; the order information is processed to obtain the processed valid order data; a navigation distance estimation model is constructed, and the navigation distance estimation model is trained using the valid order data to obtain a trained target navigation distance estimation model; the navigation distance of the current ride order is estimated using the target navigation distance estimation model. Thus, by using a linear regression model and considering the characteristics of the city, adjustments are made according to the characteristics of different cities, thereby more accurately reflecting the actual road and traffic conditions and achieving more accurate distance estimation. The present invention solves a practical and common problem in the online car-hailing industry and has high practical value.

[0042] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A schematic diagram showing a flow chart of a navigation distance estimation method provided by an embodiment of the present invention;

[0045] Figure 2 A schematic diagram showing the structure of a navigation distance estimation device provided by an embodiment of the present invention is shown;

[0046] Figure 3 A schematic diagram showing the structure of a navigation distance estimation device provided by another embodiment of the present invention is shown;

[0047] Figure 4 A schematic diagram of the physical structure of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0048] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary 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 limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0049] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0050] The embodiment of the present invention provides a navigation distance estimation method, such as Figure 1 As shown, at least the following steps S101 to S104 may be included:

[0051] Step S101, obtaining order information of multiple historical ride orders.

[0052] The historical ride orders in the embodiment of the present invention are historical data before the time node of the current ride order. The order information may include city information (IP of the city where the order is located), order time, departure coordinates, destination coordinates, and actual navigation distance, etc.

[0053] Step S102, performing data processing on the order information to obtain processed valid order data.

[0054] The embodiment of the present invention constructs a navigation distance estimation model, and estimates the actual navigation between the current location and the destination location using the navigation distance estimation model when the actual navigation distance cannot be obtained. In order to obtain sample data for training the navigation distance estimation model, the embodiment of the present invention requires historical order information for data processing.

[0055] Specifically, first, data preprocessing can be performed on the order information to obtain valid order information. Data preprocessing can include removing duplicate records, clearing missing values, and outlier processing. Removing duplicate records means checking and deleting identical records in the order information, including rows with identical field values; the specific operation of clearing missing values ​​is to clear rows containing missing values; outlier processing can include removing records with extreme values ​​of actual navigation distance exceeding a preset distance range, removing records with identical departure and destination places, and removing records with extreme values ​​of travel time exceeding a preset time range.

[0056] It should be noted that the preset distance range and the preset time range can be set according to actual needs. Preferably, the preset distance range is 500 meters to 500 kilometers, that is, records of actual navigation distances less than 500 meters or greater than 500 kilometers are eliminated; the preset time range is 1 minute to 24 hours, that is, records of abnormally short (such as less than 1 minute) or abnormally long (such as more than 24 hours) travel time are eliminated.

[0057] Furthermore, the valid order information can be grouped according to the city information to obtain multiple valid order information corresponding to each target city; the Euclidean distance between the departure coordinates and the destination coordinates in each valid order information is calculated, and the Euclidean distance is added to the valid order information; the valid order information is normalized and / or standardized to obtain valid order data. In this way, the valid order data corresponding to each city ID can be obtained, and the navigation distance estimation model can be subsequently trained using the valid order data.

[0058] Step S103, constructing a navigation distance estimation model, and using valid order data to train the navigation distance estimation model to obtain a trained target navigation distance estimation model.

[0059] Specifically, the embodiment of the present invention selects distance features and time features as model features to construct a navigation distance estimation model based on linear regression. Linear regression is a regression analysis that uses a least squares function called a linear regression equation to model the relationship between one or more independent variables and a dependent variable. The least squares function is a linear combination of one or more model parameters called regression coefficients. Compared with simple linear regression, considering that the road conditions between different cities and the road conditions at different times may be quite different, the embodiment of the present invention adds city features and time features as auxiliary features. City features are to construct separate navigation distance estimation models for different city IDs and train them based on the valid order data corresponding to the city ID. Distance features are the main features of the navigation distance estimation model, which are directly reflected in the model formula, including Euclidean distance and actual navigation distance. Time features include hour information, week information and weekend information, that is, the specific time of the trip, the day of the week and whether it is a weekend are extracted according to the order time.

[0060] The navigation distance estimation model based on linear regression in the embodiment of the present invention is expressed by the following formula (1):

[0061] y=β0+β1x+∈(1)

[0062] Wherein, y represents the actual navigation distance; x represents the Euclidean distance; β0 and β1 are model parameters respectively; ∈ is the error term. It should be noted that the model parameters in the embodiment of the present invention are estimated using the least squares method.

[0063] Furthermore, for the navigation distance estimation model of different city IDs, the corresponding valid order data is used for training and testing. Specifically, for each city ID, the corresponding valid order data is divided into a training sample set and a test sample set. Preferably, the proportion of the training sample set and the test sample set can be 80% and 20%. Specifically, the navigation distance estimation model is trained using the training sample set; the trained navigation distance estimation model is tested using the test sample set, and the navigation distance estimation model is cross-validated by calculating the model performance index, and the model hyperparameters (such as regularization strength, etc.) are adjusted according to the cross-validation results to obtain a navigation distance estimation model that meets the performance standards, which is used as the target navigation distance estimation model corresponding to each city information that has been trained. Among them, the model performance indicators may include mean square error (MSE), determination coefficient (R 2 ) and mean absolute error (MAE) to evaluate the accuracy and reliability of the model.

[0064] Finally, set the version number of the target navigation distance estimation model and save it, and record the model version, training data features, performance indicators and other metadata. In practical applications, model deployment is also required to call the target navigation distance estimation model for estimation when the actual navigation distance cannot be obtained. Model deployment includes three parts: environment preparation, model encapsulation, and load balancing and expansion. It is understandable that when preparing the environment, a dedicated production server or cloud environment needs to be set up to ensure the consistency of the production environment and the development environment, including software versions, dependent libraries, etc. Then, the model is encapsulated as an API or microservice to facilitate other system calls. Specifically, container technology (such as Docker) can be used to encapsulate the model and its dependencies to ensure environmental consistency. In the process of model estimation, the request pressure can be dispersed by configuring a load balancer. In addition, an automatic expansion mechanism can be implemented to dynamically adjust resources according to demand to ensure normal model calls and data estimation.

[0065] Step S104, estimating the navigation distance of the current ride order using the target navigation distance estimation model.

[0066] It should be noted that since the situation where the actual navigation distance cannot be obtained may occur during driving or before driving, the target navigation distance estimation model proposed in the embodiment of the present invention can use the current position coordinates as the departure coordinates, calculate the Euclidean distance between the current position coordinates and the destination coordinates, and estimate the actual navigation distance.

[0067] Furthermore, the embodiment of the present invention can also perform real-time performance monitoring on the target navigation distance estimation model, and perform abnormal warning and version update on the target navigation distance estimation model based on the model performance. The real-time performance monitoring in the embodiment of the present invention includes model performance monitoring and system health monitoring. For model performance monitoring, the performance indicators of the model can be calculated in real time, such as the root mean square error RMSE, the mean absolute error MAE and other indicators, so as to judge the prediction accuracy of the monitoring model in real time according to the indicators, and each indicator can be checked for the indicator performance threshold. When the indicator is lower than the indicator performance threshold, an alarm is triggered to remind relevant personnel to perform version update or model repair on the model. For system health monitoring, the system resources such as CPU usage, memory consumption, disk space and other system resources deployed by the monitoring model can be monitored, and the API response time and throughput can be monitored to control its operation within the normal range.

[0068] Furthermore, the embodiments of the present invention can also log and analyze the model operation. Specifically, the input, output and intermediate results of the model can be recorded in detail, and log analysis tools can be used to regularly analyze the logs to identify potential problems.

[0069] Furthermore, the embodiments of the present invention can also comprehensively evaluate the model performance according to a preset time period. For example, the model performance is comprehensively evaluated every week or month, and business goals are set to determine whether the model needs to be updated. When the model version is updated, the historical order data needs to be updated and the model needs to be retrained and tested. Specifically, the embodiments of the present invention can establish an automated process, establish an automated process, regularly collect historical order data, and process it to obtain new training data. In addition, data quality checks can also be implemented to ensure the reliability of the new data. When retraining the model with new training data, incremental learning technology can be used to update the model without losing the learned information, thereby obtaining an updated target navigation distance estimation model.

[0070] Furthermore, the embodiment of the present invention can also perform A / B testing, that is, compare the performance of the new and old models to gradually transfer traffic to the new model with better performance. For the updated target navigation distance estimation model, a version number (i.e., a unique identifier) ​​needs to be set and saved. The embodiment of the present invention records the cause and effect of each update by maintaining a detailed version history.

[0071] Furthermore, the embodiment of the present invention may also establish a feedback mechanism to collect user feedback and opinions from various business departments on navigation distance estimation, and adjust the model or optimize the prediction process based on the feedback.

[0072] In addition, in the actual experiment based on the navigation distance estimation method proposed by the present invention, the estimation method of the present invention and the prior art are compared. Taking the data of Beijing as an example (100,000 random historical orders as samples, the ratio of training data to test data is 8:2), the error rate of the present invention is reduced by 34% compared with the algorithm of Euclidean distance × 1.4; compared with the algorithm of Manhattan distance, the error rate is reduced by 28%. It can be seen that the navigation distance estimation method proposed by the present invention effectively improves the accuracy of navigation distance estimation and greatly improves the reputation and customer satisfaction of the online car-hailing platform.

[0073] The embodiment of the present invention provides a navigation distance estimation method, including obtaining order information of multiple historical ride orders; performing data processing on the order information to obtain processed valid order data; constructing a navigation distance estimation model, training the navigation distance estimation model using the valid order data to obtain a trained target navigation distance estimation model; and estimating the navigation distance of the current ride order using the target navigation distance estimation model. Through the present invention, at least the following technical effects can be achieved:

[0074] (1) Accuracy: The present invention enables the model to provide more accurate navigation distance estimates by comprehensively considering multiple features, including urban characteristics.

[0075] (2) Flexibility and scalability: The model can not only be adjusted according to the characteristics of different cities, but can also be easily extended to other application scenarios.

[0076] (3) Practicality and commercial value: The present invention solves a common problem in the online car-hailing industry and has significant commercial application value.

[0077] (4) Automation and efficiency: Through automatic data collection and model updating, the present invention can greatly improve operational efficiency.

[0078] Further, as Figure 1 The specific implementation of the present invention provides a navigation distance estimation device, such as Figure 2 As shown, the device may include: an information acquisition module 210 , a data processing module 220 , a model building module 230 and a distance estimation module 240 .

[0079] The information acquisition module 210 can be used to acquire order information of multiple historical ride orders;

[0080] The data processing module 220 can be used to process the order information to obtain processed valid order data;

[0081] The model building module 230 can be used to build a navigation distance estimation model, and train the navigation distance estimation model using valid order data to obtain a trained target navigation distance estimation model;

[0082] The distance estimation module 240 may be used to estimate the navigation distance of future ride orders using a target navigation distance estimation model.

[0083] Alternatively, if Figure 3 As shown, another embodiment of the present invention provides a navigation distance estimation device, which further includes: a performance monitoring module 250.

[0084] The performance monitoring module 250 may be used to perform real-time performance monitoring on the target navigation distance estimation model, and to perform abnormal warning and / or version update on the target navigation distance estimation model based on the model performance.

[0085] Optionally, the data processing module 220 may also be used to perform data preprocessing on the order information to obtain valid order information; the data preprocessing includes at least one of removing duplicate records, clearing missing values, and processing abnormal values;

[0086] Group the valid order information according to the city information to obtain multiple valid order information corresponding to each target city;

[0087] Calculate the Euclidean distance between the origin coordinates and the destination coordinates in each valid order information, and add the Euclidean distance to the valid order information;

[0088] The valid order information is normalized and / or standardized to obtain valid order data.

[0089] Optionally, the model building module 230 may also be used to select distance features and time features as model features to build a navigation distance estimation model based on linear regression;

[0090] Wherein, the distance feature includes Euclidean distance and / or actual navigation distance; the time feature includes hour information, week information and / or weekend information;

[0091] The navigation distance estimation model based on linear regression is expressed by the following formula:

[0092] y=β0+β1x+∈

[0093] Among them, y represents the actual navigation distance; x represents the Euclidean distance; β0 and β1 are model parameters respectively; ∈ is the error term.

[0094] Optionally, the model building module 230 may also be used to divide the corresponding valid order data into a training sample set and a test sample set for each city information;

[0095] Using the training sample set to train the navigation distance estimation model;

[0096] The trained navigation distance estimation model is tested using the test sample set, and the navigation distance estimation model is cross-validated by calculating the model performance index to obtain a navigation distance estimation model that meets the performance standard as the target navigation distance estimation model corresponding to each city information after training;

[0097] The model performance index includes at least one of mean square error, coefficient of determination and mean absolute error.

[0098] It should be noted that for other corresponding descriptions of the functional modules involved in the navigation distance estimation device provided in the embodiment of the present invention, reference can be made to Figure 1 The corresponding description of the method shown will not be repeated here.

[0099] Based on the above Figure 1 The method shown, accordingly, an embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the navigation distance estimation method described in any of the above embodiments are implemented.

[0100] Based on the above Figure 1 The method shown and Figure 3 The embodiment of the device shown in the figure, the embodiment of the present invention also provides a physical structure diagram of a computer device, such as Figure 4 As shown, the computer device may include a communication bus, a processor, a memory and a communication interface, and may also include an input / output interface and a display device, wherein each functional unit may communicate with each other through the bus. The memory stores a computer program, and the processor is used to execute the program stored in the memory and execute the steps of the navigation distance estimation method described in the above embodiment.

[0101] Those skilled in the art can clearly understand that the specific working processes of the systems, devices, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and for the sake of brevity, they are not further described here.

[0102] In addition, the functional units in various embodiments of the present invention may be physically independent of each other, or two or more functional units may be integrated together, or all functional units may be integrated into one processing unit. The above integrated functional units may be implemented in the form of hardware, or in the form of software or firmware.

[0103] Those skilled in the art can understand that if the integrated functional unit is implemented in the form of software and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can essentially or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, which includes a number of instructions to enable a computing device (such as a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention when running the instructions. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk and other media that can store program codes.

[0104] Alternatively, all or part of the steps of implementing the aforementioned method embodiments may be accomplished by hardware associated with program instructions (such as a computing device such as a personal computer, a server, or a network device), and the program instructions may be stored in a computer-readable storage medium. When the program instructions are executed by a processor of a computing device, the computing device executes all or part of the steps of the methods described in the embodiments of the present invention.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that within the spirit and principles of the present invention, the technical solutions described in the aforementioned embodiments can still be modified, or some or all of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate from the protection scope of the present invention.

Claims

1. A navigation distance estimation method, characterized in that: The method comprises: Get order information of multiple historical ride orders; Performing data processing on the order information to obtain processed valid order data; Constructing a navigation distance estimation model, and training the navigation distance estimation model using the valid order data to obtain a trained target navigation distance estimation model; The target navigation distance estimation model is used to estimate the navigation distance of the current ride order.

2. The method according to claim 1, characterized in that The order information includes at least one of city information, order time, departure coordinates, destination coordinates and actual navigation distance.

3. The method according to claim 2, characterized in that The step of processing the order information to obtain processed valid order data includes: Performing data preprocessing on the order information to obtain valid order information; the data preprocessing includes at least one of removing duplicate records, clearing missing values, and processing outliers; Grouping the valid order information according to the city information to obtain multiple pieces of valid order information corresponding to each target city; Calculate the Euclidean distance between the coordinates of the departure place and the coordinates of the destination in each valid order information, and add the Euclidean distance to the valid order information; The valid order information is normalized and / or standardized to obtain valid order data.

4. The method according to claim 3, characterized in that The constructing of the navigation distance estimation model comprises: Distance features and time features are selected as model features to build a navigation distance estimation model based on linear regression; The distance feature includes Euclidean distance and / or actual navigation distance; the time feature includes hour information, week information and / or weekend information.

5. The method according to claim 4, characterized in that The navigation distance estimation model based on linear regression is expressed by the following formula: y=β0+β1x+∈ Among them, y represents the actual navigation distance; x represents the Euclidean distance; β0 and β1 are model parameters respectively; ∈ is the error term.

6. The method according to claim 2, characterized in that The using the valid order data to train the navigation distance estimation model to obtain a trained target navigation distance estimation model includes: For each of the city information, the corresponding valid order data is divided into a training sample set and a test sample set; Using the training sample set to train the navigation distance estimation model; The trained navigation distance estimation model is tested using the test sample set, and the navigation distance estimation model is cross-validated by calculating the model performance index to obtain a navigation distance estimation model that meets the performance standard as the target navigation distance estimation model corresponding to each of the city information that has been trained; The model performance index includes at least one of mean square error, determination coefficient and mean absolute error.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: The target navigation distance estimation model is subjected to real-time performance monitoring, and abnormal warning and / or version updating are performed on the target navigation distance estimation model based on the model performance.

8. A navigation distance estimation device, characterized in that: The device comprises: An information acquisition module, used to acquire order information of multiple historical ride orders; A data processing module, used to process the order information to obtain processed valid order data; A model building module, used to build a navigation distance estimation model, and train the navigation distance estimation model using the valid order data to obtain a trained target navigation distance estimation model; The distance estimation module is used to estimate the navigation distance of future ride orders using the target navigation distance estimation model.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the navigation distance estimation method according to any one of claims 1 to 7.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the navigation distance estimation method according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Order value predicting method and equipment

    CN104599002A

  • ETA prediction, training and display method and device

    CN116263853A

  • Verifying a road network of a map

    US20150377629A1

  • Method and syetem for determining transport service fee

    US20180101877A1