Automatic driving method and device, vehicle and storage medium

Abstract

The application relates to an automatic driving method and device, a vehicle and a storage medium. The automatic driving method comprises the following steps: acquiring a learning track recorded and stored by a vehicle, acquiring track information of the learning track, navigation information corresponding to the track information and lane line information corresponding to the track information from the learning track; identifying different scenes on the learning track and determining scene intervals according to at least one of the track information, the navigation information and the lane line information or a combination thereof; and executing different automatic driving strategies according to the different scenes and the scene intervals. The scheme provided by the application can automatically identify different scenes and execute different driving strategies correspondingly, so that the vehicle can smoothly pass through various different scenes when automatically driving, the automatic driving pass rate of different scenes is improved, and the automatic driving experience of a user is improved.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to an autonomous driving method, device, vehicle and storage medium. Background Technology

[0002] Autonomous driving utilizes various sensors, computer vision, artificial intelligence, and machine learning technologies to perceive, analyze, and make decisions about the road environment, thereby enabling vehicles to navigate and control autonomously.

[0003] In the autonomous driving methods of related technologies, when vehicles pass through complex scenarios such as intersections, roundabouts, U-turns, and dedicated right-turn lanes, their limited field of vision makes it impossible to obtain scene information in a timely manner, resulting in the inability to smoothly pass through these complex scenarios and affecting the autonomous driving experience. Summary of the Invention

[0004] To address or partially address the problems existing in related technologies, this application provides an autonomous driving method, device, vehicle, and storage medium that can automatically identify different scenarios and execute different driving strategies accordingly, thereby enabling the vehicle to smoothly pass through various scenarios during autonomous driving, improving the autonomous driving pass rate in different scenarios and enhancing the user's autonomous driving experience.

[0005] The first aspect of this application provides an autonomous driving method, comprising:

[0006] The system acquires the learning trajectory recorded and stored by the vehicle, and extracts the trajectory information, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information from the learning trajectory.

[0007] Based on at least one or a combination of the trajectory information, the navigation information, and the lane line information, identify different scenes on the learning trajectory and determine scene intervals;

[0008] Different autonomous driving strategies are executed based on the different scenarios and scenario ranges.

[0009] In one embodiment, identifying different scenes on the learning trajectory based on at least one or a combination of the trajectory information, the navigation information, and the lane line information includes:

[0010] Based on the matching result between the trajectory information and the navigation information, a set scenario on the learning trajectory is identified. The set scenario includes one of the following: roundabout scenario, U-turn scenario, right-turn lane scenario, main / auxiliary road switching scenario; or,

[0011] Based on the trajectory information and the lane line information, the divergence or merging scenarios on the learned trajectory are identified; or,

[0012] Based on the lane line information, the intersection scene on the learning trajectory is identified.

[0013] In one embodiment, identifying a set scene on the learning trajectory based on the matching result of the trajectory information and the navigation information includes:

[0014] The road link segments and navigation points in the navigation information are matched with the trajectory information to obtain the positions of the link segments and navigation points on the learning trajectory;

[0015] Determine the curvature and direction of travel of the learning trajectory;

[0016] Based on the attribute information corresponding to the link segment, and / or the curvature and direction of travel of the learning trajectory, the set scene on the learning trajectory is identified.

[0017] In one embodiment, identifying the divergence or merging scenario on the learned trajectory based on the trajectory information and the lane line information includes:

[0018] The road width change is determined based on the trajectory information and the lane line information;

[0019] Based on the different changes in road width, the divergence or merging scenarios on the learning trajectory are identified.

[0020] In one embodiment, determining the road width change based on the trajectory information and the lane line information includes:

[0021] Determine the trajectory points in the trajectory information;

[0022] Determine the road boundary lines in the lane line information;

[0023] The road width is determined based on the trajectory points and the road boundary line, and the changes in road width are determined based on the road width.

[0024] In one embodiment, identifying the divergence or merging scenarios on the learning trajectory based on different road width variations includes:

[0025] Based on the change in road width from decreasing to increasing, the traffic diversion scenario on the learning trajectory is identified; or,

[0026] Based on the change in road width to increase, the merging scene on the learning trajectory is identified.

[0027] In one embodiment, identifying the intersection scene on the learning trajectory based on the lane line information includes:

[0028] Based on the stop line information in the lane line information, the intersection scene on the learning trajectory is identified.

[0029] In one embodiment, after identifying different scenes on the learning trajectory and determining scene intervals, the method further includes:

[0030] Detect the clarity and / or completeness of the lane lines at the locations of the trajectory points in the trajectory information;

[0031] Based on the clarity and / or completeness of the lane lines, and the different scenarios, the scenario range is optimized to obtain the optimized scenario range;

[0032] The execution of different autonomous driving strategies based on the different scenarios and scenario ranges includes:

[0033] Different autonomous driving strategies are executed based on the different scenarios and the optimized scenario ranges.

[0034] A second aspect of this application provides an autonomous driving device, comprising:

[0035] The acquisition module is used to acquire the learning trajectory recorded and stored by the vehicle, and to acquire the trajectory information of the learning trajectory, the navigation information corresponding to the trajectory information, and the lane line information corresponding to the trajectory information from the learning trajectory.

[0036] The scene recognition module is used to identify different scenes on the learning trajectory and determine scene intervals based on at least one or a combination of the trajectory information, the navigation information, and the lane line information.

[0037] The strategy driving module is used to execute different autonomous driving strategies based on the different scenarios and scenario ranges.

[0038] In one embodiment, the scene recognition module includes:

[0039] The first scene recognition submodule is used to identify a set scene on the learning trajectory based on the matching result of the trajectory information and the navigation information. The set scene includes one of the following: roundabout scene, U-turn scene, right-turn lane scene, main-auxiliary road switching lane scene; or,

[0040] The second scene recognition submodule is used to identify the divergence or merging scene on the learned trajectory based on the trajectory information and the lane line information; or,

[0041] The third scene recognition submodule is used to identify the intersection scene on the learning trajectory based on the lane line information.

[0042] In one embodiment, the apparatus further includes:

[0043] An interval optimization module is used to detect the clarity and / or completeness of lane lines at the locations of trajectory points in the trajectory information; based on the clarity and / or completeness of the lane lines and the different scenarios, the module optimizes the scene intervals to obtain optimized scene intervals.

[0044] The strategy driving module executes different autonomous driving strategies based on the different scenarios and the optimized scenario range.

[0045] A third aspect of this application provides a vehicle, comprising:

[0046] Processor; and

[0047] A memory that stores executable code, which, when executed by the processor, causes the processor to perform the method described above.

[0048] A fourth aspect of this application provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor of a vehicle, causes the processor to perform the method described above.

[0049] The technical solution provided in this application may include the following beneficial effects:

[0050] The technical solution of this application, after the vehicle completes the learning of the driving path and stores the learning trajectory, can obtain the trajectory information of the learning trajectory, the navigation information corresponding to the trajectory information, and the lane line information corresponding to the trajectory information from the learning trajectory; then, based on at least one or a combination of the trajectory information, the navigation information, and the lane line information, different scenarios on the learning trajectory are identified and scenario intervals are determined; in this way, different autonomous driving strategies can be executed according to the different scenarios and scenario intervals, thereby enabling the vehicle to smoothly pass through various different scenarios during autonomous driving, improving the autonomous driving pass rate in different scenarios and enhancing the user's autonomous driving experience.

[0051] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0052] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments thereof.

[0053] Figure 1 This is a flowchart illustrating the autonomous driving method according to an embodiment of this application;

[0054] Figure 2This is a flowchart illustrating an autonomous driving method according to another embodiment of this application;

[0055] Figure 3 This is a schematic diagram of the road width in the autonomous driving method shown in the embodiments of this application;

[0056] Figure 4 This is a schematic diagram illustrating the identification of a traffic diversion scenario in the autonomous driving method shown in the embodiments of this application;

[0057] Figure 5 This is a schematic diagram of the structure of an autonomous driving device shown in an embodiment of this application;

[0058] Figure 6 This is a schematic diagram of the structure of an autonomous driving device shown in another embodiment of this application;

[0059] Figure 7 This is a schematic diagram of the vehicle structure shown in the embodiments of this application. Detailed Implementation

[0060] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0061] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0062] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0063] Existing autonomous driving methods in related technologies cannot acquire scene information in complex scenarios in a timely manner, resulting in difficulties in smoothly navigating complex scenarios and impacting the autonomous driving experience. This application provides an autonomous driving method that can automatically identify different scenarios and execute different driving strategies accordingly, thereby enabling the vehicle to smoothly navigate various scenarios during autonomous driving, improving the autonomous driving pass rate in different scenarios and enhancing the user's autonomous driving experience.

[0064] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0065] Figure 1 This is a flowchart illustrating the autonomous driving method in an embodiment of this application.

[0066] See Figure 1 The method includes:

[0067] S101. Obtain the learning trajectory recorded and stored by the vehicle, and obtain the trajectory information, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information from the learning trajectory.

[0068] In this embodiment of the application, the vehicle can first enable the path learning function, record the set path traveled as the learning path, and store it after the learning is completed.

[0069] For example, a vehicle can be programmed with navigation information, such as a starting point and an ending point. It then travels according to this navigation information, recording its trajectory and various environmental information perceived by its sensors. This information is stored upon reaching the destination. The recorded and stored trajectory can be called a learning trajectory.

[0070] The learning trajectory includes recording the vehicle's driving path, which consists of a series of coordinates of trajectory points sorted by time. These trajectory points were recorded during the learning process. Subsequent scene recognition and classification of these trajectory points, through defined methods, can assist users in autonomous driving, significantly improving the vehicle's passability in various scenarios such as intersections, roundabouts, U-turns, main and auxiliary roads, and areas lacking lane markings.

[0071] S102. Based on at least one or a combination of trajectory information, navigation information, and lane line information, identify different scenes on the learning trajectory and determine scene intervals.

[0072] The identification of different scenes on the learning trajectory based on at least one or a combination of trajectory information, navigation information, and lane line information may include:

[0073] Based on the matching results of trajectory information and navigation information, the designated scenarios on the learning trajectory are identified. These scenarios include one of the following: roundabout scenario, U-turn scenario, right-turn lane scenario, main / auxiliary road switching scenario; or...

[0074] Based on trajectory information and lane line information, identify divergence or merging scenarios on the learning trajectory; or,

[0075] Based on lane line information, the intersection scene on the learning trajectory is identified.

[0076] Specifically, based on the matching results of trajectory information and navigation information, the set scene on the learning trajectory is identified, including: matching the road line link segments and navigation points in the navigation information with the trajectory information to obtain the positions of the link segments and navigation points on the learning trajectory; determining the curvature and driving direction of the learning trajectory; and identifying the set scene on the learning trajectory based on the attribute information corresponding to the link segments, and / or the curvature and driving direction of the learning trajectory.

[0077] Specifically, based on trajectory information and lane line information, the divergence or merging scenarios on the learning trajectory are identified, including: determining road width changes based on trajectory information and lane line information; and identifying divergence or merging scenarios on the learning trajectory based on different road width changes.

[0078] Among them, based on lane line information, the intersection scene on the learning trajectory is identified, including: based on the stop line information in the lane line information, the intersection scene on the learning trajectory is identified.

[0079] S103. Execute different autonomous driving strategies according to different scenarios and scenario ranges.

[0080] Once different scenarios and scenario ranges are determined, vehicles can use corresponding specific strategies to plan routes and execute different autonomous driving strategies in different scenarios, thereby significantly improving the vehicle's passability in various scenarios and enhancing the autonomous driving experience.

[0081] The technical solution of this application embodiment, after the vehicle completes the learning of the driving path and stores the learning trajectory, can obtain the trajectory information of the learning trajectory, the navigation information corresponding to the trajectory information, and the lane line information corresponding to the trajectory information from the learning trajectory; then, based on at least one or a combination of the trajectory information, navigation information, and lane line information, different scenarios on the learning trajectory are identified and scenario intervals are determined; in this way, different autonomous driving strategies can be executed according to different scenarios and scenario intervals, thereby enabling the vehicle to smoothly pass through various different scenarios during autonomous driving, improving the autonomous driving pass rate of different scenarios and improving the user's autonomous driving experience.

[0082] Figure 2This is a flowchart illustrating an autonomous driving method according to another embodiment of this application.

[0083] The technical solution of this application embodiment is based on the learning trajectory that has been learned and recorded. It can analyze road information in advance, thereby identifying different scenarios in advance and obtaining the accurate interval position of the scenario. This allows different driving strategies to be executed according to different scenarios and scenario intervals.

[0084] See Figure 2 The method includes:

[0085] S201. Obtain the learning trajectory recorded and stored by the vehicle, and obtain the trajectory information of the learning trajectory, the navigation information corresponding to the trajectory information, and the lane line information corresponding to the trajectory information from the learning trajectory.

[0086] In this embodiment of the application, the vehicle can first enable the path learning function, record the set path traveled as the learning path, and store it after the learning is completed.

[0087] For example, a vehicle can be programmed with navigation information, such as a starting point and an ending point. It then travels according to this navigation information, recording its trajectory and various environmental information perceived by its sensors. This information is stored upon reaching the destination. The recorded and stored trajectory can be called a learning trajectory.

[0088] In other words, the learning trajectory recorded and stored after the vehicle completes path learning can include trajectory information of the learning trajectory, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information.

[0089] When a vehicle is learning a route, navigation can be activated. Starting from the beginning of the route, the vehicle will follow the navigation route to the end of the route. During the journey, the vehicle can continuously record the navigation information, the trajectory information of the route, and the lane line information perceived by the vehicle's sensors.

[0090] The learning trajectory includes recording the vehicle's driving path, which consists of a series of coordinates of trajectory points sorted by time. These trajectory points were recorded during the learning process. Subsequent scene recognition and classification of these trajectory points, through defined methods, can assist users in autonomous driving, significantly improving the vehicle's passability in various scenarios such as intersections, roundabouts, U-turns, main and auxiliary roads, and areas lacking lane markings.

[0091] It should be noted that the learning trajectory in this application embodiment can be the driving trajectory in the vehicle commuting mode. The commuting mode in this application embodiment should be broadly understood as the driving mode used by the user on their weekday commuting routes, and the driving mode used by the user on high-frequency driving routes between certain frequently used locations on weekends or holidays, and is not limited to weekday travel. For example, if a user frequently travels between their residence and a fixed entertainment venue on weekends, the driving trajectory with the user's residence and the aforementioned entertainment venue as the starting and ending points is also applicable to the above-mentioned commuting mode. Autonomous driving in commuting mode can also be called AI-assisted driving. Through one learning session, a learning trajectory can be recorded and generated, and a set number of learning trajectories can be stored. For example, a user can store 5 or 10 learning trajectories in the vehicle or in the cloud. The route length of each learning trajectory can be a preset length, for example, up to 100 kilometers, but not limited to this.

[0092] In this application embodiment, the acquisition of trajectory information of the learning trajectory, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information can be performed in the following ways, but is not limited to these:

[0093] In some examples, the vehicle can be programmed with navigation information, such as a starting point and an ending point, and then drive according to that information. As the vehicle travels from the starting point to the ending point along the navigation path, it continuously acquires road information and vehicle-related information through various sensors installed on the vehicle. This includes information such as vehicle speed, lane markings, traffic sign placement, traffic light locations, landmark locations, and information about intersections, roundabouts, U-turn lanes, dedicated right-turn lanes, main and auxiliary roads, and lanes before and after toll booths—all determined by these traffic elements. This information is temporarily saved as it is acquired. When the vehicle reaches the ending point or loses power while in park, the vehicle determines the end of the current trip, and the information acquisition process concludes. The vehicle can then store all acquired driving information, including the learned path's trajectory information, the corresponding navigation information, and the corresponding lane marking information, on the vehicle's end or in the cloud.

[0094] S202. Based on the matching results of trajectory information and navigation information, identify the set scene on the learning trajectory and determine the scene interval.

[0095] Based on the matching results of trajectory information and navigation information, the set scenarios on the learning trajectory are identified. The set scenarios include one of the following: roundabout scenario, U-turn scenario, right turn lane scenario, and main-auxiliary road switching scenario.

[0096] For example, the road link segments and navigation points in the navigation information are matched with the trajectory information to obtain the positions of the link segments and navigation points on the learning trajectory; based on the attribute information corresponding to the link segments, the set scene on the learning trajectory is identified.

[0097] There is usually an error between the latitude and longitude coordinates of the learning trajectory and the navigation information. You can first trim the path information in the navigation information to get the navigation link (road line) segment corresponding to the learning trajectory, and then combine it with the navigation points to map the navigation information to the specific coordinate points of the trajectory in the learning trajectory.

[0098] In navigation systems, a link is the basic unit of a road model and the smallest unit of navigation data, referring to a road segment from one intersection to the next. The two ends of a link are the intersection nodes on the map. A node represents a virtual node object in the road network, which can be approximated as an intersection on a real road. One node is the starting node, and the other is the ending node. A link is a curved object representing a path between nodes, consisting of two nodes and several shape points. Generally, in map data, nodes and links can directly represent the entire road topology.

[0099] In this embodiment, the link segment of the navigation information is matched with the trajectory in the learning trajectory using map matching. Map matching combines user location information and map data to calculate the user's accurate location on the road on the map, assisting in the precise control of the in-vehicle navigation. Navigation points are points where turns are required during navigation. For example, if you want to turn left 50 meters ahead, the location 50 meters ahead is the navigation point. By matching the trajectory information with the navigation information, the trajectory positions of the navigation link and navigation points on the learning trajectory can be determined. Then, the link and navigation point can be mapped onto the trajectory of the learning trajectory.

[0100] The link segments in the navigation information contain attribute information, which records the corresponding scenario (special scenario) for that link segment. For example, it may record whether the link segment corresponds to a roundabout scenario, a U-turn scenario, a right-turn lane scenario, or a main-auxiliary road switching scenario. Therefore, based on the attribute information corresponding to the link segment, the scenario on the learning trajectory can be identified as a roundabout scenario, a U-turn scenario, a right-turn lane scenario, or a main-auxiliary road switching scenario.

[0101] Furthermore, based on the position information of the matched link segment in the learning trajectory, the scene range of the set scenario can be roughly obtained, such as the scene range of a roundabout scenario, a U-turn scenario, a right-turn lane scenario, or a main-auxiliary road switching scenario. For example, if the total length of the learning trajectory is 1000 meters, the scene range of the right-turn lane is between 500 meters and 520 meters.

[0102] In addition, by combining the curvature of the learning trajectory and the driving direction, the set scenarios on the learning trajectory can be identified. In this way, the set scenarios can be identified and the approximate range of different scenarios can be determined under the dual verification of navigation information and trajectory trend.

[0103] For example, in a U-turn scenario, the trajectory follows a 180-degree turn. The curvature and direction angle can be calculated using the first derivative of the trajectory, thereby determining where the U-turn scenario interval of the learned trajectory begins and ends.

[0104] S203. Based on the trajectory information and lane line information, identify the divergence or merging scenarios on the learning trajectory and determine the scenario intervals.

[0105] Specifically, road width changes can be determined based on trajectory information and lane line information; and based on different road width changes, divergence or merging scenarios on the learning trajectory can be identified.

[0106] Determining road width changes based on trajectory information and lane line information can include:

[0107] Determine the trajectory points in the trajectory information; determine the road boundary lines in the lane line information; determine the road width based on the trajectory points and the road boundary lines; and determine the changes in road width based on the road width.

[0108] Among these, identifying divergence or merging scenarios on the learning trajectory based on different road width variations can include:

[0109] Based on the change in road width from decreasing to increasing, the traffic diversion scenarios on the learning trajectory are identified; or,

[0110] Based on the change in road width, the merging scene on the learning trajectory is identified.

[0111] In other words, it can examine the road boundary lines (also known as curbs) to the left and right of the coordinate points of the learning trajectory, calculate the road width on both sides based on the trajectory points and the road boundary lines, and identify the divergence or merging scenarios on the learning trajectory based on the different changes in road width.

[0112] Furthermore, by extracting location information where road width changes, the scene range of diversion or merging scenarios can be determined.

[0113] The following methods can be used to determine road width, but are not limited to these:

[0114] See Figure 3 Given the lane line information around trajectory point P, rays can be emitted perpendicularly from trajectory point P to the left and right road boundary lines L1 and L2 respectively, until the rays intersect with the road boundary lines L1 and L2 (curb). At this point, the starting point of the ray is trajectory point P, and the intersection points are P1 with the left road boundary line L1 and P2 with the right road boundary line L2. The distance from the starting point to the intersection point is the left and right road widths, respectively. That is, the distance S1 between trajectory point P and intersection point P1 is the left road width, and the distance S2 between trajectory point P and intersection point P2 is the right road width. The road width is the sum of the left and right road widths, i.e., (S1 + S2).

[0115] See Figure 4 If the road width becomes S3, and S3 is less than (S1+S2), that is, the road width becomes smaller, it is identified as a diversion scenario; if the road width suddenly becomes larger, it is identified as a merging scenario.

[0116] It should also be noted that the diversion or merging scenarios can be identified by combining changes in the number of lanes, or by using machine learning models to identify diversion or merging scenarios. Machine learning models can utilize features such as guide lines, main roads and auxiliary roads.

[0117] During model training, trajectory features can be derived from big data filtering, and the training trajectories can be pre-filtered to remove illegal driving trajectories before being used to train the model for recognition. Trajectory features can include information such as the trajectory's angle and curvature.

[0118] S204. Based on lane line information, identify intersection scenes on the learning trajectory and determine scene intervals.

[0119] Specifically, the system can identify intersection scenarios on the learning trajectory based on the stop line information within the lane line information. In other words, it can determine whether a scenario is an intersection based on the stop line information within the lane line information; for example, a scenario with a stop line in a non-roundabout area is identified as an intersection scenario.

[0120] Furthermore, based on the location of the intersection in the learning path, the scene range of the intersection scene can be determined.

[0121] It should also be noted that the embodiments of this application can also use the Occupancy Network (OCC) machine learning model to visually identify the intersection scene. OCC is a deep learning-based 3D environment perception method that uses neural networks to predict whether a spatial location is occupied. In real road environments, OCC can be used to identify obstacles that have not been marked or trained, such as temporary obstacles, construction areas, tree branches extending from the roadside, and goods that have fallen off a truck ahead.

[0122] Furthermore, there is no necessary sequential relationship between steps S202, S203, and S204; they can be executed separately.

[0123] S205. Detect the clarity and / or completeness of the lane lines at the locations of the trajectory points in the trajectory information; optimize the scene intervals based on the clarity and / or completeness of the lane lines and different scenarios to obtain optimized scene intervals.

[0124] When identifying different scenarios and determining the corresponding scenario intervals in the aforementioned steps, the accuracy of the initially determined scenario intervals may be relatively poor. Therefore, embodiments of this application can further optimize the scenario intervals. For example, this application can detect whether the lane lines at the location of the trajectory are clear based on the lane lines around the trajectory, and detect whether the lane lines around the trajectory coordinate points are intact, and detect whether the direction and trend of the lane lines are consistent with the trajectory. Thus, based on the clarity and / or integrity of the lane lines, a more accurate lane line interval can be determined, and it can be determined where there are lane lines and where there are no lane lines.

[0125] In this step, based on the clarity and / or completeness of the lane lines, the previously determined scene range can be further fine-tuned using lane line ranges, thereby making the scene range obtained after optimization more accurate. For example, this application combines the feature information of navigation information, trajectory information, and lane line information, and uses the clarity and / or completeness of lane lines as an aid to obtain a more accurate scene range, such as a scene range accurate to within 0.5 meters.

[0126] S206. Execute different autonomous driving strategies based on different scenarios and optimized scenario ranges.

[0127] Once the scene range is determined, or a more precise scene range is determined through optimization, the vehicle can use corresponding specific strategies to plan routes in different scenes, thereby significantly improving the vehicle's passability in various scenarios and enhancing the autonomous driving experience.

[0128] For example, in some scenarios, it is acceptable to drive along the lane lines, such as on ordinary roads; in other scenarios, it is not appropriate to drive along the lane lines, such as in diverging or merging scenarios where you need to change lanes immediately, or at intersections where there are no lane lines and you cannot follow the lane lines, or in U-turn scenarios or some right-turn lane scenarios where it is better not to drive along the lane lines.

[0129] For example, for a roundabout scenario, the corresponding autonomous driving strategy could be to remain in the roundabout without changing lanes; for a right-turn lane scenario, the corresponding autonomous driving strategy could be to enter and exit on the right; and for a U-turn scenario, the corresponding autonomous driving strategy could be to use a planned path generated by a planner with high curvature.

[0130] The method in this application embodiment can combine feature information such as navigation information, trajectory information, and lane line information to identify different scenarios on the learning trajectory and determine scenario intervals. For example, it can identify roundabout scenarios, U-turn scenarios, right-turn lane scenarios, and main-auxiliary road switching scenarios. This allows the vehicle to adopt different driving strategies for different scenarios when it is driving autonomously in different scenarios, based on the identified scenarios and corresponding scenario intervals. This enables the vehicle to smoothly pass through various scenarios during autonomous driving, improving the autonomous driving pass rate in different scenarios and enhancing the user's autonomous driving experience.

[0131] Corresponding to the aforementioned application function implementation method embodiments, this application also provides an autonomous driving device, a vehicle, and corresponding embodiments.

[0132] Figure 5 This is a schematic diagram of the structure of an autonomous driving device shown in an embodiment of this application.

[0133] See Figure 5 An autonomous driving device 50 includes: an acquisition module 51, a scene recognition module 52, and a strategy driving module 53.

[0134] The acquisition module 51 is used to acquire the learning trajectory recorded and stored by the vehicle, and to obtain trajectory information, corresponding navigation information, and lane line information from the learning trajectory. In this embodiment, the vehicle can first activate the path learning function, record the set path traveled as the learning path, and store it after learning is complete. For example, the vehicle can set navigation information, such as a path start and end point, and then drive according to the navigation information, recording the vehicle's trajectory while driving, and recording various environmental information perceived by the vehicle's various sensors, storing it after reaching the path end. The trajectory recorded and stored by the vehicle can be called the learning trajectory.

[0135] The scene recognition module 52 is used to identify different scenes on the learning trajectory and determine scene intervals based on at least one or a combination of trajectory information, navigation information, and lane line information. For example, the scene recognition module 52 identifies a set scene on the learning trajectory based on the matching result of trajectory information and navigation information. The set scene includes one of the following: roundabout scene, U-turn scene, right-turn lane scene, main-auxiliary road switching lane scene; or, based on trajectory information and lane line information, it identifies a divergence scene or merging scene on the learning trajectory; or, based on lane line information, it identifies an intersection scene on the learning trajectory.

[0136] The strategy driving module 53 is used to execute different autonomous driving strategies based on different scenarios and scenario ranges. Once the different scenarios and scenario ranges are determined, the strategy driving module 53 can use corresponding specific strategies to plan routes and execute different autonomous driving strategies in different scenarios, thereby significantly improving the vehicle's passability in various scenarios and enhancing the autonomous driving experience.

[0137] The autonomous driving device provided in this application, after the vehicle completes the learning of the driving path and stores the learning trajectory, can obtain the trajectory information of the learning trajectory, the navigation information corresponding to the trajectory information, and the lane line information corresponding to the trajectory information from the learning trajectory; then, based on at least one or a combination of the trajectory information, navigation information, and lane line information, it can identify different scenarios on the learning trajectory and determine the scenario intervals; in this way, different autonomous driving strategies can be executed according to different scenarios and scenario intervals, thereby enabling the vehicle to smoothly pass through various different scenarios during autonomous driving, improving the autonomous driving pass rate of different scenarios and improving the user's autonomous driving experience.

[0138] Figure 6 This is a schematic diagram of the structure of an autonomous driving device shown in another embodiment of this application.

[0139] See Figure 6 An autonomous driving device 50 includes: an acquisition module 51, a scene recognition module 52, a strategy driving module 53, and a section optimization module 54.

[0140] The functions of the acquisition module 51, scene recognition module 52, and strategy driving module 53 can be found in [reference needed]. Figure 5 As described in the text, it will not be repeated here.

[0141] The scene recognition module 52 may include: a first scene recognition submodule 521, a second scene recognition submodule 522, or a third scene recognition submodule 523.

[0142] The first scene recognition submodule 521 is used to identify the set scene on the learning trajectory based on the matching result of trajectory information and navigation information. The set scene includes one of the following: roundabout scene, U-turn scene, right turn lane scene, and main and auxiliary road switching lane scene.

[0143] The second scene recognition submodule 522 is used to identify the divergence or merging scene on the learning trajectory based on the trajectory information and lane line information.

[0144] The third scene recognition submodule 523 is used to identify intersection scenes on the learning trajectory based on lane line information.

[0145] The first scene recognition submodule 521 matches the road link segments and navigation points in the navigation information with the trajectory information to obtain the positions of the link segments and navigation points on the learning trajectory; determines the curvature and driving direction of the learning trajectory; and identifies the set scene on the learning trajectory based on the attribute information corresponding to the link segments and / or the curvature and driving direction of the learning trajectory.

[0146] The second scene recognition submodule 522 determines the road width change based on the trajectory information and lane line information; and identifies the diverging or merging scenarios on the learning trajectory according to different road width changes. For example, it identifies the diverging scenario on the learning trajectory based on the road width change being smaller; or it identifies the merging scenario on the learning trajectory based on the road width change being larger.

[0147] Among them, the third scene recognition submodule 523 identifies the intersection scene on the learning trajectory based on the stop line information in the lane line information.

[0148] The interval optimization module 54 is used to detect the clarity and / or completeness of the lane lines at the location of the trajectory points in the trajectory information; based on the clarity and / or completeness of the lane lines and different scenarios, the scene interval is optimized to obtain the optimized scene interval;

[0149] The strategy driving module 53 executes different autonomous driving strategies based on different scenarios and optimized scenario ranges.

[0150] The autonomous driving device shown in this application embodiment can combine navigation information, trajectory information, lane line information and other feature information to identify different scenarios on the learning trajectory and determine scenario intervals. For example, it can identify roundabout scenarios, U-turn scenarios, right-turn lane scenarios, main and auxiliary road switching scenarios, etc. This allows the vehicle to adopt different driving strategies for different scenarios when driving autonomously in different scenarios, based on the identified different scenarios and corresponding scenario intervals. This enables the vehicle to smoothly pass through various scenarios during autonomous driving, improving the autonomous driving pass rate in different scenarios and enhancing the user's autonomous driving experience.

[0151] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0152] Figure 7 This is a schematic diagram of the vehicle structure shown in the embodiments of this application.

[0153] See Figure 7 The vehicle 1000 includes a memory 1010 and a processor 1020.

[0154] The processor 1020 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0155] Memory 1010 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by processor 1020 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 1010 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory 1010 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital multifunction optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, a high-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0156] The memory 1010 stores executable code, which, when processed by the processor 1020, can cause the processor 1020 to execute part or all of the methods described above.

[0157] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0158] Alternatively, this application may also be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of a vehicle (or electronic device, or server, etc.), causes the processor to perform part or all of the steps of the above-described method according to this application.

[0159] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An autonomous driving method, characterized in that, include: The system acquires the learning trajectory recorded and stored by the vehicle, and extracts trajectory information, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information from the learning trajectory; wherein the trajectory information includes trajectory points recorded by the vehicle while it is learning the route. Based on at least one or a combination of the trajectory information, the navigation information, and the lane line information, identify different scenes on the learning trajectory and determine scene intervals; Different autonomous driving strategies are executed based on the different scenarios and scenario ranges. The method of identifying different scenes on the learning trajectory based on at least one or a combination of the trajectory information, the navigation information, and the lane line information includes: Based on the matching result between the trajectory information and the navigation information, a set scenario on the learning trajectory is identified. The set scenario includes one of the following: roundabout scenario, U-turn scenario, right-turn lane scenario, main / auxiliary road switching scenario; or, Based on the trajectory information and the lane line information, the divergence or merging scenarios on the learned trajectory are identified; or, Based on the lane line information, the intersection scene on the learning trajectory is identified.

2. The method according to claim 1, characterized in that, The step of identifying a set scene on the learning trajectory based on the matching result of the trajectory information and the navigation information includes: The road link segments and navigation points in the navigation information are matched with the trajectory information to obtain the positions of the link segments and navigation points on the learning trajectory; Determine the curvature and direction of travel of the learning trajectory; Based on the attribute information corresponding to the link segment, and / or the curvature and direction of travel of the learning trajectory, the set scene on the learning trajectory is identified.

3. The method according to claim 1, characterized in that, The step of identifying the divergence or merging scenarios on the learned trajectory based on the trajectory information and the lane line information includes: The road width change is determined based on the trajectory information and the lane line information; Based on the different changes in road width, the divergence or merging scenarios on the learning trajectory are identified.

4. The method according to claim 3, characterized in that, The step of determining the road width change based on the trajectory information and the lane line information includes: Determine the trajectory points in the trajectory information; Determine the road boundary lines in the lane line information; The road width is determined based on the trajectory points and the road boundary line, and the changes in road width are determined based on the road width.

5. The method according to claim 3, characterized in that, The process of identifying diverging or merging scenarios on the learning trajectory based on different road width variations includes: Based on the change in road width from decreasing to increasing, the traffic diversion scenario on the learning trajectory is identified; or, Based on the change in road width to increase, the merging scene on the learning trajectory is identified.

6. The method according to claim 1, characterized in that, The step of identifying the intersection scene on the learning trajectory based on the lane line information includes: Based on the stop line information in the lane line information, the intersection scene on the learning trajectory is identified.

7. The method according to any one of claims 1 to 6, characterized in that, After identifying different scenes on the learning trajectory and determining the scene intervals, the process further includes: Detect the clarity and / or completeness of the lane lines at the locations of the trajectory points in the trajectory information; Based on the clarity and / or completeness of the lane lines, and the different scenarios, the scenario range is optimized to obtain the optimized scenario range; The execution of different autonomous driving strategies based on the different scenarios and scenario ranges includes: Different autonomous driving strategies are executed based on the different scenarios and the optimized scenario ranges.

8. An autonomous driving device, characterized in that, include: The acquisition module is used to acquire the learning trajectory recorded and stored by the vehicle, and to acquire trajectory information, navigation information corresponding to the trajectory information, and lane line information corresponding to the trajectory information from the learning trajectory; wherein the trajectory information includes trajectory points recorded by the vehicle while learning the route; The scene recognition module is used to identify different scenes on the learning trajectory and determine scene intervals based on at least one or a combination of the trajectory information, the navigation information, and the lane line information. The strategy driving module is used to execute different autonomous driving strategies according to the different scenarios and scenario ranges; The scene recognition module includes: The first scene recognition submodule is used to identify a set scene on the learning trajectory based on the matching result of the trajectory information and the navigation information. The set scene includes one of the following: roundabout scene, U-turn scene, right-turn lane scene, main-auxiliary road switching lane scene; or, The second scene recognition submodule is used to identify the divergence or merging scene on the learned trajectory based on the trajectory information and the lane line information; or, The third scene recognition submodule is used to identify the intersection scene on the learning trajectory based on the lane line information.

9. The apparatus according to claim 8, characterized in that, The device further includes: An interval optimization module is used to detect the clarity and / or completeness of lane lines at the locations of trajectory points in the trajectory information; based on the clarity and / or completeness of the lane lines and the different scenarios, the module optimizes the scene intervals to obtain optimized scene intervals. The strategy driving module executes different autonomous driving strategies based on the different scenarios and the optimized scenario range.

10. A vehicle, characterized in that, include: processor; as well as A memory having executable code stored thereon, which, when executed by the processor, causes the processor to perform the method as described in any one of claims 1-7.

11. A computer-readable storage medium, characterized in that: It stores executable code that, when executed by the vehicle's processor, causes the processor to perform the method as described in any one of claims 1-7.

Images