Personalized navigation method, navigation large model training method, device and equipment

Abstract

The invention relates to a personalized navigation method, a navigation large model training method, a device and equipment. According to the invention, the driving habit of the user is obtained by performing deep learning on the historical travel route data, the real-time traffic data and the historical driving behavior data of the user, and the route planning is further performed according to the destination input by the user based on the driving habit. Therefore, the driving habit of the user and the real-time traffic condition can be fully considered in the route planning process to obtain the candidate travel route, more comprehensive travel suggestions are provided for the user, personalized customization of the travel route is realized, and the navigation flexibility is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of navigation technology, and in particular to a personalized navigation method, a navigation large model training method, a device and equipment. Background Art

[0002] Navigation is a common means of assisted driving. Map navigation can provide users with route planning, driving reminders and other services.

[0003] Currently, a common navigation method is to obtain a route starting point and a route end point input by a user, and search for multiple optional routes from the starting point to the end point based on map data.

[0004] However, in the above navigation method, the route planned based on the same starting point and end point is relatively simple, which makes the navigation flexibility relatively poor. Summary of the Invention

[0005] In order to solve the above technical problems, the present disclosure provides a personalized navigation method, a navigation large model training method, an apparatus and a device to improve the flexibility of the navigation method.

[0006] In a first aspect, an embodiment of the present disclosure provides a personalized navigation method, including:

[0007] Conduct deep learning on historical travel route data, real-time traffic data, and users' historical driving behavior data to obtain users' driving preference data;

[0008] Route planning is performed based on the current route data and map data input by the user to obtain multiple basic travel routes;

[0009] The multiple basic travel routes are calibrated according to the user's driving preference data to obtain multiple candidate travel routes.

[0010] In some embodiments, the deep learning of historical travel route data, real-time traffic data, and historical driving behavior data of the user to obtain the user's driving preference data includes:

[0011] Each path point contained in the historical travel route data, the attribute category of each path point, real-time traffic data, and the user's historical driving behavior data are input into the trained deep learning driving behavior data model to obtain the driving preference data output by the trained deep learning driving behavior data model.

[0012] In some embodiments, the current route data includes at least one waypoint, and the route planning is performed based on the current route data input by the user and the map data to obtain multiple basic travel routes, including:

[0013] Obtain the attribute category of each path point based on map data retrieval;

[0014] Each of the path points and the attribute category of each of the path points are input into a trained travel route retrieval model to obtain a plurality of basic travel routes output by the trained travel route retrieval model.

[0015] In some embodiments, the multiple basic travel routes are calibrated according to the user's driving preference data to obtain multiple candidate travel routes, including:

[0016] The multiple basic travel routes and the driving preference data are input into a trained travel route recommendation model to obtain multiple candidate travel routes output by the trained travel route recommendation model.

[0017] In some embodiments, the method further comprises:

[0018] Presenting the plurality of candidate travel routes to the user;

[0019] selecting a target travel route from the plurality of candidate travel routes according to a user's selection operation on the plurality of candidate travel routes;

[0020] Navigating based on the target travel route;

[0021] The driving preference data and the historical travel route data are updated according to the target travel route and the user's current driving behavior data for the target travel route.

[0022] In some embodiments, the waypoints in the current route data include a starting point, and navigating based on the target travel route includes:

[0023] Updating the starting point in the current route data to the current position of the vehicle in real time;

[0024] updating the target travel route based on the updated current route data to obtain an updated target travel route;

[0025] In response to a route switching instruction from the user, navigation is performed based on the updated target travel route.

[0026] In some embodiments, the driving preference data includes at least one or more of the following:

[0027] Driving speed, parking habits, acceleration and deceleration habits.

[0028] In a second aspect, an embodiment of the present disclosure provides a method for training a large navigation model. The large navigation model is applied to the personalized navigation method described in the first aspect. The large navigation model includes at least a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model. The method includes:

[0029] Based on each path point included in the historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as inputs of the travel route retrieval model, the historical travel routes in the historical travel route data are used as outputs of the travel route retrieval model, and the travel route retrieval model is trained to obtain a trained travel route retrieval model;

[0030] Inputting each path point included in the historical travel route data, the attribute category of each path point, real-time traffic data, and the user's historical driving behavior data into a deep learning driving behavior data model, so that the deep learning driving behavior data model learns the driver's driving preference data to obtain a trained deep learning driving behavior data model;

[0031] Using the historical travel route data and driving preference data as inputs to the travel route recommendation model, and using the historical travel route data as outputs of the travel route recommendation model to train the travel route recommendation model, thereby obtaining a trained travel route recommendation model;

[0032] The navigation model is updated and iterated according to the target travel route data and the user's current driving behavior data for the target travel route.

[0033] In some embodiments, updating and iterating the navigation model based on the target travel route data and the user's current driving behavior data for the target travel route includes:

[0034] Adding the target route data corresponding to the target travel route to the historical travel route data to obtain updated historical travel route data;

[0035] Adding the current driving behavior data to the historical driving behavior data to obtain updated historical driving behavior data;

[0036] Based on each path point included in the updated historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as inputs of the trained travel route retrieval model, the historical travel routes in the updated historical travel routes are used as outputs of the trained travel route retrieval model, and the trained travel route retrieval model is updated to obtain an updated travel route retrieval model;

[0037] Inputting each path point included in the updated historical travel route data, the attribute category of each path point, the real-time traffic data, and the updated historical driving behavior data into the trained deep learning driving behavior data model, so that the trained deep learning driving behavior data model learns the updated driver's driving preference data, thereby obtaining an updated deep learning driving behavior data model;

[0038] The updated historical travel route data and the updated driver's driving preference data are used as inputs of the trained travel route recommendation model, and the updated historical travel route data is used as outputs of the trained travel route recommendation model to update the trained travel route recommendation model to obtain an updated travel route recommendation model.

[0039] In a third aspect, an embodiment of the present disclosure provides a personalized navigation device, including:

[0040] The learning module is used to conduct in-depth learning on historical travel route data, real-time traffic data, and the user's historical driving behavior data to obtain the user's driving preference data;

[0041] The planning module is used to plan routes based on the current route data and map data input by the user to obtain multiple basic travel routes;

[0042] The calibration module is used to calibrate the multiple basic travel routes according to the user's driving preference data to obtain multiple candidate travel routes.

[0043] In a fourth aspect, an embodiment of the present disclosure provides a navigation large model training device, comprising:

[0044] a first training module for training the travel route retrieval model based on each path point, the attribute category of each path point, the historical travel route, and real-time traffic data contained in the historical travel route data as inputs to the travel route retrieval model, and using the historical travel routes in the historical travel route data as outputs of the travel route retrieval model to obtain a trained travel route retrieval model;

[0045] a second training module, configured to input each waypoint included in the historical travel route data, the attribute category of each waypoint, real-time traffic data, and the user's historical driving behavior data into a deep learning driving behavior data model, so that the deep learning driving behavior data model learns the driver's driving preference data, thereby obtaining a trained deep learning driving behavior data model;

[0046] a third training module, configured to use the historical travel route data and the driving preference data as inputs of the travel route recommendation model, and train the travel route recommendation model using the historical travel route data as outputs of the travel route recommendation model to obtain a trained travel route recommendation model;

[0047] The second updating module is used to iterate the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route.

[0048] In a fifth aspect, an embodiment of the present disclosure provides an electronic device, including:

[0049] Memory;

[0050] processor; and

[0051] computer programs;

[0052] The computer program is stored in the memory and is configured to be executed by the processor to implement the method as described in the first aspect and / or the second aspect.

[0053] In a sixth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the method described in the first aspect and / or the second aspect.

[0054] In a seventh aspect, an embodiment of the present disclosure provides a vehicle comprising the device, electronic device or computer-readable storage medium as described above.

[0055] The personalized navigation method, navigation large model training method, device and equipment provided by the embodiments of the present disclosure obtain the user's driving habits through deep learning of historical travel route data, real-time traffic data and the user's historical driving behavior data, and further plan the route according to the destination input by the user based on the driving habits. In this way, the user's driving habits and real-time traffic conditions can be fully considered in the route planning process to obtain candidate travel routes, provide users with more comprehensive travel suggestions, realize personalized customization of travel routes, and improve the flexibility of navigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0057] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0058] Figure 1 A flowchart of a personalized navigation method provided by an embodiment of the present disclosure;

[0059] Figure 2 A schematic diagram of an application scenario provided by an embodiment of the present disclosure;

[0060] Figure 3 A flowchart of a navigation model training method provided in an embodiment of the present disclosure;

[0061] Figure 4 A schematic diagram of a personalized navigation method provided by another embodiment of the present disclosure;

[0062] Figure 5 A schematic diagram of the structure of a personalized navigation device provided by an embodiment of the present disclosure;

[0063] Figure 6 A schematic structural diagram of a navigation large model training device provided by an embodiment of the present disclosure;

[0064] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0065] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0066] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0067] The embodiments of the present disclosure provide a personalized navigation method, which is described below in conjunction with specific embodiments.

[0068] Figure 1 This is a flow chart of the personalized navigation method provided by the embodiment of the present disclosure. The method can be applied to Figure 2In the illustrated application scenario, the application scenario includes a vehicle 21 and a server 22, wherein the vehicle 21 is equipped with an onboard device, which may specifically be a vehicle computer, a smartphone, a PDA, a tablet computer, a laptop computer, an all-in-one computer, an intelligent driving device, etc. It is understood that the personalized navigation method provided in the embodiments of the present disclosure can also be applied in other scenarios.

[0069] Below Figure 1 The personalized navigation method shown in FIG. 1 is introduced, and the specific steps of the method are as follows:

[0070] S101. Perform deep learning on historical travel route data, real-time traffic data, and historical driving behavior data of the user to obtain the user's driving preference data.

[0071] Specifically, the historical travel route data includes at least a plurality of historical travel routes, attribute categories of waypoints in the historical travel routes, and the like.

[0072] Real-time traffic data includes but is not limited to the actual traffic volume, road congestion, road restrictions, etc. obtained in the current period and predicted traffic volume in the future period.

[0073] Alternatively, real-time traffic data can be obtained through Internet of Vehicles technology, which connects vehicles with other vehicles, traffic facilities, and traffic management departments, enabling the sharing and acquisition of real-time traffic information.

[0074] A user's historical driving behavior data includes, but is not limited to: control data, including acceleration, braking, parking, gear position, high and low beam / fog lights / position lights, windows, seat belts, steering wheel angle, steering wheel speed, etc.; usage data, including mileage, life cycle, driving section, driving time, driving direction, driving time, driving frequency, single driving duration, congestion duration, and unimpeded driving duration; performance data, including accelerator pedal opening, brake pedal percentage, brake pedal status, engine speed, instantaneous fuel consumption, fuel consumption per 100 kilometers, and remaining fuel level. Driving behavior data helps us understand users' driving habits and behavior patterns, thereby providing more personalized navigation services.

[0075] In some embodiments, a user's driving preference data can be acquired through a trained deep learning driving behavior data model. The deep learning driving behavior data model is fed with waypoints and their attribute categories in historical travel routes, the user's historical driving behavior data, and real-time traffic data corresponding to the historical travel routes. This allows the deep learning driving behavior data model to learn the correlation between the user's historical driving behavior and other data, thereby enabling the trained deep learning driving behavior data model to output the user's driving preferences for different waypoints and their attribute categories, different travel routes, and different traffic conditions, i.e., the user's driving preference data.

[0076] S102: performing route planning based on the current route data and map data input by the user to obtain multiple basic travel routes.

[0077] The current route data includes at least one waypoint, i.e., the route and destination of the user's trip. Specifically, the waypoint may include one or more of the starting point, end point, and waypoint input by the user.

[0078] Optionally, the current route data includes at least the destination of the trip, that is, the end point.

[0079] Optionally, when the user does not input a starting point, the user's current location or the vehicle's current location may be used as the starting point for route planning.

[0080] The user may provide the current route data by inputting the current route data through voice or text, or by selecting a point on a map, etc., and this embodiment of the present disclosure does not limit this.

[0081] The current route data entered by the user is matched with the map data to determine the actual geographical location corresponding to the current route data. Path planning is further performed in combination with the road network information in the map data to obtain multiple basic travel routes. This process can use any path planning algorithm, such as the Dijkstra algorithm, the A* algorithm, etc.

[0082] Specifically, the basic travel route is a route from a starting point to an end point; if the user inputs any one or more waypoints, the basic travel route should pass through each waypoint.

[0083] S103: Calibrate the multiple basic travel routes according to the user's driving preference data to obtain multiple candidate travel routes.

[0084] The user's driving preference data refers to the user's driving habits, such as driving speed, parking habits, acceleration and deceleration habits, etc.

[0085] Optionally, a corresponding driving preference database is established for each user. When the user performs navigation operations, the current user identity is obtained based on the currently logged-in account and / or identity authentication operations such as face recognition, and the driving preference database corresponding to the user is matched based on the user identity to obtain the user's driving preference data.

[0086] In some embodiments, multiple basic travel routes are calibrated based on the user's driving preference data, including adjusting route planning based on the basic travel routes, such as helping users avoid congested roads, giving priority to expressways or ordinary roads based on the driving speed in the user's driving preference data, helping users find parking lots around the destination in advance based on the user's parking habits, etc.

[0087] In some embodiments, calibrating multiple basic travel routes based on the user's driving preference data also includes adjusting the priority order of the multiple basic travel routes. For example, the multiple basic travel routes may be arranged in descending order of driving distance. After calibrating the multiple basic travel routes, the priority of the route that better suits the driver's driving habits may be increased.

[0088] The disclosed embodiment obtains the user's driving preference data by performing deep learning on historical travel route data, real-time traffic data, and the user's historical driving behavior data; performs route planning based on the current route data and map data input by the user to obtain multiple basic travel routes; calibrates the multiple basic travel routes based on the user's driving preference data to obtain multiple candidate travel routes, obtains the user's driving habits by performing deep learning on the historical travel route data, real-time traffic data, and the user's historical driving behavior data, and further performs route planning based on the user's input destination based on the driving habits, thereby fully considering the user's driving habits and real-time traffic conditions in the route planning process to obtain candidate travel routes, providing the user with more comprehensive travel suggestions, realizing personalized customization of travel routes, and improving the flexibility of navigation.

[0089] Based on the above embodiment, the method also includes: presenting the multiple candidate travel routes to the user; selecting a target travel route from the multiple candidate travel routes based on the user's selection operation for the multiple candidate travel routes; navigating based on the target travel route; and updating the driving preference data and historical travel route data based on the target travel route and the user's current driving behavior data for the target travel route.

[0090] Optionally, if there is a historical travel route that has been selected by the user among the multiple candidate travel routes, the historical travel route is marked to remind the user that the historical travel route has been selected.

[0091] Optionally, the user can select the target travel route through voice, text input, or by clicking on the interface.

[0092] Optionally, after the user completes the selection, the target travel route is marked and is displayed first during the next navigation of the same route.

[0093] Optionally, the in-vehicle device guides the user to travel along the target travel route through voice prompts, text prompts or image displays.

[0094] The disclosed embodiments provide an intuitive and friendly user interaction interface, enabling users to conveniently interact with and provide feedback to the in-vehicle device system.

[0095] In some embodiments, the personalized navigation method may be implemented based on a navigation macro model. Figure 3 A flowchart of a navigation model training method provided by an embodiment of the present disclosure. The following is a navigation model training method, wherein the navigation model at least includes a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model. Figure 3 As shown, the method includes:

[0096] S301. Based on each path point contained in the historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as the input of the travel route retrieval model, and the historical travel routes in the historical travel route data as the output of the travel route retrieval model, the travel route retrieval model is trained to obtain a trained travel route retrieval model.

[0097] S302. Input each path point contained in the historical travel route data, the attribute category of each path point, real-time traffic data, and the user's historical driving behavior data into a deep learning driving behavior data model, so that the deep learning driving behavior data model learns the driver's driving preference data and obtains a trained deep learning driving behavior data model.

[0098] S303: Using the historical travel route data and driving preference data as inputs to the travel route recommendation model, and using the historical travel route data as outputs of the travel route recommendation model to train the travel route recommendation model, thereby obtaining a trained travel route recommendation model.

[0099] S304: updating and iterating the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route.

[0100] Specifically, updating and iterating the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route includes:

[0101] The target route data corresponding to the target travel route is added to the historical travel route data to obtain updated historical travel route data; the current driving behavior data is added to the historical driving behavior data to obtain updated historical driving behavior data; each path point contained in the updated historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data are used as inputs of the trained travel route retrieval model, and the historical travel routes in the updated historical travel routes are used as outputs of the trained travel route retrieval model to update the trained travel route retrieval model to obtain an updated travel route retrieval model; the updated Each path point contained in the historical travel route data, the attribute category of each path point, the real-time traffic data, and the updated historical driving behavior data are input into the trained deep learning driving behavior data model, so that the trained deep learning driving behavior data model learns the updated driver's driving preference data to obtain an updated deep learning driving behavior data model; the updated historical travel route data and the updated driver's driving preference data are used as inputs to the trained travel route recommendation model, and the updated historical travel route data is used as outputs of the trained travel route recommendation model to update the trained travel route recommendation model to obtain an updated travel route recommendation model.

[0102] The application of the above navigation model will be described in detail below with reference to specific embodiments.

[0103] In some embodiments, the current route data includes at least one path point, and the route planning is performed based on the current route data input by the user and the map data to obtain multiple basic travel routes, including: obtaining the attribute category of each path point based on the map data; inputting each path point and the attribute category of each path point into a trained travel route retrieval model to obtain multiple basic travel routes output by the trained travel route retrieval model.

[0104] The map includes multiple points of interest (POIs), which can represent any location data relevant to the user, such as businesses, public venues, and transportation hubs. This information typically includes the name, address, phone number, business hours, etc., and related geographic location information can be searched by POI coordinates or keywords.

[0105] The attribute categories of POI points cover all aspects of people's daily lives, such as: catering services: including restaurants, cafes, bars, etc.; shopping services: including shopping malls, supermarkets, convenience stores, etc.; life services: including banks, post offices, hospitals, etc.; sports and leisure services: including parks, gyms, cinemas, etc.; medical care services: including hospitals, clinics, pharmacies, etc.; accommodation services: including hotels, B&Bs, etc.; scenic spots: including tourist attractions, parks, etc.

[0106] The waypoints in the current route data correspond to the points of interest in the map data. Therefore, the attribute categories of the corresponding waypoints can be determined based on the attribute categories of the points of interest. Based on the attribute categories of the waypoints, the in-vehicle device can better identify the user's intentions and provide personalized services.

[0107] Specifically, basic travel routes can be planned using a trained travel route retrieval model. First, historical travel data is obtained, and the attribute categories of path points in historical travel routes are determined based on the attribute categories of each POI point in the map data (e.g., scenic area, service area, downtown area, etc.). Furthermore, the path points in the historical travel routes, their attribute categories, and the historical travel route data are input into the travel route retrieval model, allowing the travel route retrieval model to learn the relevant features between the historical travel route data, the path points in the historical travel routes, and their attribute categories. The travel route retrieval model is then trained to obtain a trained travel route retrieval model.

[0108] Optionally, the path points in the historical travel routes and the attribute categories of the path points in the historical travel routes are used as inputs of the travel route retrieval model, and the historical travel route data are used as outputs of the travel route retrieval model. The travel route retrieval model is trained to obtain a trained travel route retrieval model.

[0109] During navigation, the trained travel route retrieval model is used to obtain basic travel routes. This includes inputting the attribute category of each waypoint in the current route data into the trained travel route retrieval model, and obtaining the output of the trained travel route retrieval model, which is multiple basic travel routes.

[0110] The embodiment of the present disclosure introduces attribute categories of waypoints when planning a basic travel route, which can better identify the user's travel intention and make the planned basic travel route more in line with user needs.

[0111] In some embodiments, the multiple basic travel routes are calibrated according to the user's driving preference data and real-time traffic data to obtain multiple candidate travel routes, including: inputting each path point contained in the current route data, the attribute category of each path point, and the user's historical driving behavior data into a trained deep learning driving behavior data model to obtain driving preference data output by the trained deep learning driving behavior data model; and inputting the multiple basic travel routes and the driving preference data into a trained travel route recommendation model to obtain multiple candidate travel routes output by the trained travel route recommendation model.

[0112] Specifically, the calibration of the basic travel routes can be achieved through the trained travel route recommendation model. In some embodiments, multiple basic travel routes are calibrated according to the user's driving preference data and real-time traffic data, including adjusting the route planning based on the basic travel routes, such as helping users avoid congested sections, giving priority to expressways or ordinary roads according to the driving speed in the user's driving preference data, helping users find parking lots around the destination in advance according to the user's parking habits, etc. In some embodiments, multiple basic travel routes are calibrated according to the user's driving preference data and real-time traffic data, and also include adjusting the priority order of the multiple basic travel routes. For example, multiple basic travel routes are arranged in descending order according to the driving distance. After the calibration of the multiple basic travel routes, the priority of the route that is more in line with the driver's driving habits is increased.

[0113] The disclosed embodiments accurately capture the personalized driving habits of each user, comprehensively consider multiple factors such as the user's driving behavior, geographic location, real-time traffic, etc., and provide customized travel routes for them, thereby improving the flexibility of navigation methods.

[0114] Based on the above embodiments, the travel route retrieval model, deep learning driving behavior data model, and travel route recommendation model can be used as separate models or as sub-models in a large navigation model, and the embodiments of the present disclosure do not limit this.

[0115] In some embodiments, the method further includes: after selecting a target travel route from the multiple candidate travel routes based on the user's selection operation on the multiple candidate travel routes, the method further includes: updating the driving preference data and historical travel route data based on the target travel route.

[0116] After the user selects the target travel route, he or she can drive along the target travel route according to the prompts of the on-board equipment. At the same time, the target travel route is selected by the user from multiple candidate travel routes, which means that the target travel route is the route that best meets the current road conditions and user intentions among the multiple candidate travel routes. Therefore, the target travel route is used as a historical travel route, and the user's driving behavior data and real-time traffic data are collected during the driving process along the target travel route to update the user's driving preference data, participate in the training of the above-mentioned navigation model or travel route retrieval model, deep learning driving behavior data model, and travel route recommendation model, thereby updating the navigation model or travel route retrieval model, deep learning driving behavior data model, and travel route recommendation model. The specific training process and update process are shown in the following figure. Figure 3 The embodiment shown will not be described in detail here.

[0117] Optionally, the target travel route, the user's driving behavior data along the target travel route, and real-time traffic data are uploaded to the server for use in model training.

[0118] In some embodiments, the path points in the current route data include a starting point, and navigation is performed based on the target travel route, including: updating the starting point in the current route data to the current position of the vehicle in real time; updating the target travel route based on the updated current route data to obtain an updated target travel route; and navigating based on the updated target travel route in response to a user's route switching instruction.

[0119] When the user is traveling along the target travel route, the starting point position in the current route data is updated in real time according to the current position of the vehicle.

[0120] Furthermore, the target travel route is updated based on the updated current route data to obtain an updated target travel route. This specifically includes: performing route planning based on the updated current route data and map data to obtain multiple updated basic travel routes; and calibrating the multiple updated basic travel routes based on the user's driving preference data and real-time traffic data to obtain multiple updated candidate travel routes. The specific implementation process is similar to the process of obtaining candidate travel routes based on the current route data in the above embodiment and is not further described here.

[0121] Specifically, the disclosed embodiments plan and recommend the optimal route for the remaining distance in real time while the user is traveling along a target travel route. For example, a user's route switching instruction determines whether to change the target travel route currently being traveled and selects the updated target travel route from the updated candidate travel routes, where the target travel route is the route from the vehicle's current location to the destination.

[0122] The disclosed embodiment ensures the real-time and accuracy of the model and navigation route by updating various data used for model training in real time, and updating the target travel route in real time based on the remaining distance while driving along the target travel route, thereby further improving the flexibility of navigation.

[0123] Figure 4 This is a schematic diagram of a personalized navigation method provided by another embodiment of the present disclosure. Figure 4 As shown, the personalized navigation method provided by the embodiment of the present disclosure uses historical travel route data (including multiple path points such as starting point, end point, and passing points) and the user's historical driving behavior data (including driving speed, parking habits, etc.) as data sources, and combines map data with traffic data training to obtain a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model.

[0124] During the navigation process, the current route data (including multiple path points such as the starting point, the end point, and the waypoints) input by the user is obtained, and the travel route retrieval model performs route planning based on the current route data input by the user and the map data to obtain multiple basic travel routes; the multiple basic travel routes and the driving preference data are input into a trained travel route recommendation model to obtain multiple candidate travel routes output by the trained travel route recommendation model; based on the user's selection operation on the multiple candidate travel routes, a target travel route is selected from the multiple candidate travel routes; and navigation is performed based on the target travel route.

[0125] At the same time, in the process of navigating based on the target travel route, the user's driving behavior data and real-time traffic data are collected in real time, the driving preference data and historical travel route data are updated, and the updated historical travel route data is used to participate in the training of the travel route retrieval model, the deep learning driving behavior data model and the travel route recommendation model, and the parameters of the travel route retrieval model, the deep learning driving behavior data model and the travel route recommendation model are updated.

[0126] In addition, during navigation based on the target travel route, the starting point in the current route data is updated in real time to the current position of the vehicle, and the target travel route is updated based on the updated current route data to obtain an updated target travel route, providing the user with updated route options.

[0127] The disclosed embodiments fully consider the user's driving habits and real-time traffic conditions during route planning to generate candidate routes. The final target route is then selected from the candidate routes based on the user's intent, thereby providing users with more comprehensive travel recommendations, enabling personalized customization of the target route, and improving navigation flexibility. By updating various data used for model training in real time and updating the target route based on the remaining distance during driving along the target route, the real-time performance and accuracy of the model and navigation route are ensured, further enhancing navigation flexibility.

[0128] Figure 5 This is a schematic diagram of the structure of the personalized navigation device provided by the embodiment of the present disclosure. The personalized navigation device can be the vehicle-mounted device as described in the above embodiment, or the personalized navigation device can be a component or assembly in the vehicle-mounted device. The personalized navigation device provided by the embodiment of the present disclosure can execute the processing flow provided by the embodiment of the personalized navigation method, such as Figure 5 As shown, the personalized navigation device 50 includes: a learning module 51, a planning module 52, and a calibration module 53; the learning module 51 is used to perform in-depth learning on historical travel route data, real-time traffic data, and the user's historical driving behavior data to obtain the user's driving preference data; the planning module 52 is used to perform route planning based on the current route data and map data input by the user to obtain multiple basic travel routes; the calibration module 53 is used to calibrate the multiple basic travel routes based on the user's driving preference data to obtain multiple candidate travel routes.

[0129] Optionally, the learning module 51 is specifically used to input each path point contained in the historical travel route data, the attribute category of each path point, the real-time traffic data, and the user's historical driving behavior data into a trained deep learning driving behavior data model to obtain the driving preference data output by the trained deep learning driving behavior data model.

[0130] Optionally, the current route data includes at least one path point, and the planning module 52 includes a retrieval unit 521 and a first acquisition unit 522; the retrieval unit 521 is used to obtain the attribute category of each path point based on the map data; the first acquisition unit 522 is used to input each path point and the attribute category of each path point into a trained travel route retrieval model to obtain multiple basic travel routes output by the trained travel route retrieval model.

[0131] Optionally, the calibration module 53 is specifically configured to input the plurality of basic travel routes and the driving preference data into a trained travel route recommendation model to obtain a plurality of candidate travel routes output by the trained travel route recommendation model.

[0132] Optionally, the personalized navigation device 50 also includes a display module 54, a selection module 55, a navigation module 56, and a first update module 57; the display module 54 is used to display the multiple candidate travel routes to the user; the selection module 55 is used to select a target travel route from the multiple candidate travel routes based on the user's selection operation for the multiple candidate travel routes; the navigation module 56 is used to navigate based on the target travel route; the first update module 57 is used to update the driving preference data and historical travel route data based on the target travel route and the user's current driving behavior data for the target travel route.

[0133] Optionally, the navigation module 56 includes a first update unit 561, a second update unit 562, and a navigation unit 563; the first update unit 561 is used to update the starting point in the current route data to the current position of the vehicle in real time; the second update unit 562 is used to update the target travel route based on the updated current route data to obtain an updated target travel route; the navigation unit 563 is used to respond to the user's route switching instruction and navigate based on the updated target travel route.

[0134] Optionally, the driving preference data includes at least one or more of the following: driving speed, parking habits, acceleration and deceleration habits.

[0135] Figure 5 The personalized navigation device of the illustrated embodiment can be used to implement the technical solution of the above method embodiment. Its implementation principle and technical effects are similar and will not be described in detail here.

[0136] Figure 6 This is a structural diagram of a navigation model training device provided in an embodiment of the present disclosure. The navigation model training device can be the vehicle-mounted device described in the above embodiment, or the navigation model training device can be a component or assembly in the vehicle-mounted device. The navigation model at least includes a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model. Figure 6As shown, the navigation model training device 60 includes: a first training module 61, a second training module 62, a third training module 63, and a second updating module 64; the first training module 61 is used to use each path point contained in the historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as the input of the travel route retrieval model, and use the historical travel routes in the historical travel route data as the output of the travel route retrieval model to train the travel route retrieval model to obtain a trained travel route retrieval model; the second training module 62 is used to use each path point contained in the historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as the input of the travel route retrieval model , real-time traffic data, and the user's historical driving behavior data are input into the deep learning driving behavior data model, so that the deep learning driving behavior data model can learn the driver's driving preference data and obtain a trained deep learning driving behavior data model; the third training module 63 is used to use the historical travel route data and driving preference data as the input of the travel route recommendation model, and use the historical travel route data as the output of the travel route recommendation model to train the travel route recommendation model to obtain a trained travel route recommendation model; the second update module 64 is used to update and iterate the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route.

[0137] Optionally, the second updating module 64 is specifically configured to add the target route data corresponding to the target travel route to the historical travel route data to obtain updated historical travel route data; add the current driving behavior data to the historical driving behavior data to obtain updated historical driving behavior data; use each path point, the attribute category of each path point, the historical travel route, and the real-time traffic data contained in the updated historical travel route data as inputs to the trained travel route retrieval model, use the historical travel routes in the updated historical travel routes as outputs of the trained travel route retrieval model, and update the trained travel route retrieval model to obtain an updated travel route retrieval model; Each path point contained in the updated historical travel route data, the attribute category of each path point, the real-time traffic data, and the updated historical driving behavior data are input into the trained deep learning driving behavior data model, so that the trained deep learning driving behavior data model learns the updated driver's driving preference data to obtain an updated deep learning driving behavior data model; the updated historical travel route data and the updated driver's driving preference data are used as inputs to the trained travel route recommendation model, and the updated historical travel route data is used as the output of the trained travel route recommendation model to update the trained travel route recommendation model to obtain an updated travel route recommendation model.

[0138] Figure 6 The navigation large model training device of the illustrated embodiment can be used to implement the technical solution of the above-mentioned method embodiment. Its implementation principle and technical effects are similar and will not be repeated here.

[0139] In addition, an embodiment of the present disclosure further provides a vehicle, which includes the personalized navigation device and / or the navigation large model training device as described in the above embodiments.

[0140] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the embodiment of the present disclosure. The electronic device may be the vehicle-mounted device described in the above embodiment. The electronic device provided by the embodiment of the present disclosure may execute the processing flow provided by the embodiment of the personalized navigation method and / or the navigation large model training method, such as Figure 7 As shown, the electronic device 70 includes: a memory 71, a processor 72, a computer program and a communication interface 73; wherein the computer program is stored in the memory 71 and is configured so that the processor 72 executes the personalized navigation method as described above.

[0141] In addition, an embodiment of the present disclosure further provides a computer-readable storage medium on which a computer program is stored. The computer program is executed by a processor to implement the personalized navigation method and / or navigation large model training method described in the above embodiments.

[0142] In addition, an embodiment of the present disclosure also provides a computer program product, which includes a computer program or instructions, and when the computer program or instructions are executed by a processor, implements the personalized navigation method and / or navigation large model training method as described above.

[0143] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including, but not limited to, object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0144] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0145] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0146] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A personalized navigation method, characterized in that: The method comprises: Conduct deep learning on historical travel route data, real-time traffic data, and users' historical driving behavior data to obtain users' driving preference data; Route planning is performed based on the current route data and map data input by the user to obtain multiple basic travel routes; The multiple basic travel routes are calibrated according to the user's driving preference data to obtain multiple candidate travel routes.

2. The method according to claim 1, characterized in that The deep learning of historical travel route data, real-time traffic data, and historical driving behavior data of the user is performed to obtain the user's driving preference data, including: Each path point contained in the historical travel route data, the attribute category of each path point, real-time traffic data, and the user's historical driving behavior data are input into the trained deep learning driving behavior data model to obtain the driving preference data output by the trained deep learning driving behavior data model.

3. The method according to claim 1, characterized in that The current route data includes at least one waypoint, and the route planning is performed based on the current route data input by the user and the map data to obtain multiple basic travel routes, including: Obtain the attribute category of each path point based on map data retrieval; Each of the path points and the attribute category of each of the path points are input into a trained travel route retrieval model to obtain a plurality of basic travel routes output by the trained travel route retrieval model.

4. The method according to claim 1, wherein The plurality of basic travel routes are calibrated according to the user's driving preference data to obtain a plurality of candidate travel routes, including: The multiple basic travel routes and the driving preference data are input into a trained travel route recommendation model to obtain multiple candidate travel routes output by the trained travel route recommendation model.

5. The method according to claim 1, wherein The method further comprises: Presenting the plurality of candidate travel routes to the user; selecting a target travel route from the plurality of candidate travel routes according to a user's selection operation on the plurality of candidate travel routes; Navigating based on the target travel route; The driving preference data and the historical travel route data are updated according to the target travel route and the user's current driving behavior data for the target travel route.

6. The method according to claim 5, characterized in that The waypoints in the current route data include a starting point, and navigating based on the target travel route includes: Updating the starting point in the current route data to the current position of the vehicle in real time; updating the target travel route based on the updated current route data to obtain an updated target travel route; In response to a route switching instruction from the user, navigation is performed based on the updated target travel route.

7. The method according to claim 1, characterized in that The driving preference data includes at least one or more of the following: Driving speed, parking habits, acceleration and deceleration habits.

8. A navigation model training method, characterized in that: The navigation model is applied to the personalized navigation method according to any one of claims 1 to 7, wherein the navigation model comprises at least a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model, and the method comprises: Based on each path point included in the historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as inputs of the travel route retrieval model, the historical travel routes in the historical travel route data are used as outputs of the travel route retrieval model, and the travel route retrieval model is trained to obtain a trained travel route retrieval model; Inputting each path point included in the historical travel route data, the attribute category of each path point, real-time traffic data, and the user's historical driving behavior data into a deep learning driving behavior data model, so that the deep learning driving behavior data model learns the driver's driving preference data to obtain a trained deep learning driving behavior data model; Using the historical travel route data and driving preference data as inputs to the travel route recommendation model, and using the historical travel route data as outputs of the travel route recommendation model to train the travel route recommendation model, thereby obtaining a trained travel route recommendation model; The navigation model is updated and iterated according to the target travel route data and the user's current driving behavior data for the target travel route.

9. The method according to claim 8, characterized in that The updating and iterating of the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route includes: Adding the target route data corresponding to the target travel route to the historical travel route data to obtain updated historical travel route data; Adding the current driving behavior data to the historical driving behavior data to obtain updated historical driving behavior data; Based on each path point included in the updated historical travel route data, the attribute category of each path point, the historical travel route, and the real-time traffic data as inputs of the trained travel route retrieval model, the historical travel routes in the updated historical travel routes are used as outputs of the trained travel route retrieval model, and the trained travel route retrieval model is updated to obtain an updated travel route retrieval model; Inputting each path point included in the updated historical travel route data, the attribute category of each path point, the real-time traffic data, and the updated historical driving behavior data into the trained deep learning driving behavior data model, so that the trained deep learning driving behavior data model learns the updated driver's driving preference data, thereby obtaining an updated deep learning driving behavior data model; The updated historical travel route data and the updated driver's driving preference data are used as inputs of the trained travel route recommendation model, and the updated historical travel route data is used as outputs of the trained travel route recommendation model to update the trained travel route recommendation model to obtain an updated travel route recommendation model.

10. A personalized navigation device, characterized in that: The device comprises: The learning module is used to conduct in-depth learning on historical travel route data, real-time traffic data, and the user's historical driving behavior data to obtain the user's driving preference data; The planning module is used to plan routes based on the current route data and map data input by the user to obtain multiple basic travel routes; The calibration module is used to calibrate the multiple basic travel routes according to the user's driving preference data to obtain multiple candidate travel routes.

11. A navigation model training device, characterized in that: The navigation model includes at least a travel route retrieval model, a deep learning driving behavior data model, and a travel route recommendation model. The device includes: a first training module for training the travel route retrieval model based on each path point, the attribute category of each path point, the historical travel route, and real-time traffic data contained in the historical travel route data as inputs to the travel route retrieval model, and using the historical travel routes in the historical travel route data as outputs of the travel route retrieval model to obtain a trained travel route retrieval model; a second training module, configured to input each waypoint included in the historical travel route data, the attribute category of each waypoint, real-time traffic data, and the user's historical driving behavior data into a deep learning driving behavior data model, so that the deep learning driving behavior data model learns the driver's driving preference data, thereby obtaining a trained deep learning driving behavior data model; a third training module, configured to use the historical travel route data and the driving preference data as inputs of the travel route recommendation model, and train the travel route recommendation model using the historical travel route data as outputs of the travel route recommendation model to obtain a trained travel route recommendation model; The second updating module is used to iterate the navigation model according to the target travel route data and the user's current driving behavior data for the target travel route.

12. An electronic device, characterized in that: include: Memory; processor; as well as computer programs; The computer program is stored in the memory and configured to be executed by the processor to implement the method according to any one of claims 1 to 7 or 8 to 9.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 or 8 to 9 is implemented.

14. A vehicle comprising: The apparatus according to claim 10 or 11; or the electronic device according to claim 12; Or, the computer-readable storage medium of claim 13.

Images