Scene library construction method, device and equipment for cut-in scene, medium and product

Abstract

This application relates to a method, apparatus, device, medium, and product for constructing a scenario library for entry scenarios. The method includes: filtering vehicle behavior within a target road segment based on raw sensing data collected by multiple roadside sensors within that segment to determine conflicting vehicle sets and lane-changing vehicle sets; the raw sensing data consists of driving information of multiple vehicles within the target road segment during their journey; determining entry vehicle information for vehicles with entry scenarios within the target road segment based on the conflicting vehicle sets and lane-changing vehicle sets; and constructing a scenario library for entry scenarios based on the entry vehicle information. This method can improve the quality of the scenario library for entry scenarios.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a method, apparatus, device, medium and product for constructing a scenario library for entry into a scenario. Background Technology

[0002] With the rapid development of autonomous driving technology, a large number of vehicles with different levels of autonomous driving have been launched on the market. Before being launched, autonomous vehicles typically use data from various driving scenarios to train their autonomous driving algorithms, in order to learn to make accurate decisions in different situations, thereby improving driving safety and reliability.

[0003] Taking the "cutting in" scenario as an example, a cutting in scenario refers to a situation in road traffic where a vehicle accelerates from an adjacent lane to cut in front of the main vehicle. If an autonomous vehicle cannot accurately predict or recognize this cutting in scenario, a rear-end collision is likely to occur. By building a scenario library for cutting in scenarios, autonomous driving algorithms can learn path planning capabilities for these scenarios, enabling autonomous vehicles to make reasonable decisions when faced with cutting in real-world driving situations.

[0004] However, the scenario library built in the relevant technologies is of poor quality, which leads to frequent failures of autonomous vehicles after they are put on the road, and urgently needs to be improved. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, device, medium, and product for constructing a scenario library for entry into scenarios, which can improve the quality of the scenario library for entry into scenarios, in order to address the above-mentioned technical problems.

[0006] In a first aspect, embodiments of this application provide a method for building a scene library for entering a scene, including:

[0007] Based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0008] Based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut in within the target road segment;

[0009] Based on the vehicle information, a scenario library for entry scenarios is built.

[0010] In one embodiment, based on raw perception data collected by various roadside sensors within the target road segment, vehicle behavior within the target road segment is filtered to determine the set of conflicting vehicles, including:

[0011] Based on the current and previous center point coordinates of each vehicle in the original sensing data, multiple candidate conflict vehicle pairs are filtered.

[0012] The conflict time of each candidate vehicle pair is determined based on the center point coordinates of each vehicle in the current time.

[0013] The set of conflicting vehicles is determined based on the conflict time of each candidate conflicting vehicle pair.

[0014] In one embodiment, multiple candidate conflict vehicle pairs are filtered based on the current and previous center point coordinates of each vehicle in the original sensing data, including:

[0015] For any two adjacent vehicles, determine the distance between their current center points based on their current center point coordinates; and determine the distance between their previous center points based on their previous center point coordinates.

[0016] If the distance between the center points at the current time is less than the distance between the center points at the previous time, then the two adjacent vehicles are identified as a candidate conflict vehicle pair.

[0017] In one embodiment, the conflict time of each candidate vehicle pair is determined based on the center point coordinates of each vehicle in the multiple candidate vehicle pairs at the current time, including:

[0018] For any candidate conflict vehicle pair, obtain the coordinates of the intersection point of the trajectories of the two conflicting vehicles in the candidate conflict vehicle pair;

[0019] Based on the coordinates of the intersection point of the trajectories of the two conflicting vehicles and the coordinates of the center points of the two conflicting vehicles, determine the distance between the coordinates of the center points of the two conflicting vehicles and the coordinates of the intersection point of their trajectories.

[0020] The collision time of the candidate vehicle pair is determined based on the distance between the center point coordinates and the trajectory intersection coordinates of the two colliding vehicles, and the speed of the two colliding vehicles at the current moment.

[0021] In one embodiment, determining the set of conflicting vehicles based on the conflict time of each candidate conflicting vehicle pair includes:

[0022] Obtain the current distance between the center points of each candidate conflict vehicle pair;

[0023] Identify candidate conflict vehicle pairs whose current center point distance is less than a preset distance threshold and whose conflict time is less than a preset time threshold;

[0024] The set of conflicting vehicles is determined based on candidate conflicting vehicle pairs whose current center point distance is less than a preset distance threshold and whose conflict time is less than a preset time threshold.

[0025] In one embodiment, based on raw perception data collected by various roadside sensors within the target road segment, vehicle behavior within the target road segment is filtered to determine a set of lane-changing vehicles, including:

[0026] Based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map, the lane number information of each vehicle in each frame is determined.

[0027] The set of vehicles changing lanes is determined based on the lane number information of each vehicle in each frame.

[0028] In one embodiment, the set of vehicles changing lanes is determined based on the lane number information of each vehicle in each frame, including:

[0029] Based on the lane number information of each vehicle in each frame, obtain the candidate vehicles whose lane number information in the current frame is different from that in the previous frame.

[0030] The set of lane-changing vehicles is determined based on the candidate vehicles whose lane number information in the current frame is different from that in the previous frame.

[0031] In one embodiment, based on the conflict vehicle set and the lane-changing vehicle set, information on vehicles cutting into the target road segment is determined, including:

[0032] For any pair of conflicting vehicles in the set of conflicting vehicles, determine the first and second conflicting vehicles based on the coordinates of the center points of the two conflicting vehicles in the pair.

[0033] If the vehicle in the previous conflict is within the set of vehicles changing lanes and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then it is determined that the conflicting vehicle pair has an entry scenario.

[0034] The lane-changing time of the vehicle that collides with the vehicle in front, the vehicle identification of the vehicle that collides with the vehicle in front, and the vehicle identification of the vehicle that collides with the vehicle behind are identified as the entry vehicle information.

[0035] In one embodiment, a scenario library for entry scenarios is built based on the entry vehicle information, including:

[0036] Based on the vehicle information, obtain the scene start timestamp and scene end timestamp;

[0037] Based on the scene start timestamp and scene end timestamp, obtain the target trajectory data and target weather data corresponding to the vehicles entering the scene; the target trajectory data and target weather data corresponding to each vehicle entering the scene refer to all vehicle trajectory data and weather data from the preset time before the start of the scene to the preset time after the end of the scene.

[0038] A scenario library is built based on target trajectory data, target weather data, and information on vehicles entering the area.

[0039] Secondly, embodiments of this application also provide a scene library building device for entering a scene, including:

[0040] The vehicle filtering module is used to filter vehicle behavior within the target road segment based on raw perception data collected by various roadside sensors within the target road segment, and to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0041] The entry scenario filtering module is used to determine the entry vehicle information within the target road segment range that has an entry scenario based on the conflict vehicle set and the lane-changing vehicle set.

[0042] The scenario library building module is used to build a scenario library for the entry scenario based on the entry vehicle information.

[0043] Thirdly, embodiments of this application also provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method in any of the embodiments of the first aspect described above.

[0044] Fourthly, embodiments of this application also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method in any of the embodiments of the first aspect described above.

[0045] Fifthly, embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the method in any of the embodiments of the first aspect described above.

[0046] The scenario library construction method, apparatus, device, medium, and product for entry scenarios provided in this application firstly filters vehicle behavior within the target road segment based on raw perception data collected by various roadside sensors within the target road segment, determining conflict vehicle sets and lane-changing vehicle sets. Then, based on the conflict vehicle sets and lane-changing vehicle sets, it determines entry vehicle information for entry scenarios within the target road segment. Finally, based on the entry vehicle information, it builds a scenario library for entry scenarios. The raw perception data consists of driving information of multiple vehicles within the target road segment during their journey. In this method, when building the scenario library for entry scenarios, conflicting vehicles and lane-changing vehicles are first screened out based on vehicle behavior within the target road segment, resulting in sets of conflicting vehicles and lane-changing vehicles. Then, the sets of conflicting vehicles and lane-changing vehicles are combined to screen for entry vehicles that exist in entry scenarios. This identifies and filters vehicles that engage in lane-changing and cutting-in behaviors and exhibit dangerous conflict behaviors as entry vehicles with existing entry scenarios, improving the accuracy of entry scenario screening and thus improving the data quality of entry vehicle information. Consequently, the quality of the constructed scenario library for entry scenarios is enhanced. Furthermore, the perception data is collected through roadside sensors, making the perception data more comprehensive. This results in a more comprehensive and complete data library for entry scenarios built based on comprehensive perception data, further improving the quality of the entry scenario library. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is an application environment diagram of a scene library building method for entering a scene in one embodiment;

[0049] Figure 2 This is a flowchart illustrating a method for building a scene library for entering a scene in one embodiment;

[0050] Figure 3 This is a schematic diagram of the process for determining the set of conflicting vehicles in one embodiment;

[0051] Figure 4 This is a flowchart illustrating the process of determining the set of lane-changing vehicles in one embodiment;

[0052] Figure 5 This is a schematic diagram of the process for determining the vehicle information to be accessed in one embodiment;

[0053] Figure 6 This is a flowchart illustrating the process of building a scenario library for entering scenarios in one embodiment;

[0054] Figure 7 This is a flowchart illustrating the method for building a scene library for entering a scene in another embodiment;

[0055] Figure 8 This is a schematic diagram of the process of scene selection and scene library construction in one embodiment;

[0056] Figure 9 This is a schematic diagram of a scene library building device for entering a scene in one embodiment;

[0057] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] The technical background of the embodiments of this application will be described below.

[0060] In recent years, autonomous driving technology has developed rapidly, and a large number of vehicles with different levels of autonomous driving have been launched. However, accidents involving autonomous vehicles occur frequently. This is ultimately due to the incomplete autonomous driving scenario library, which limits the iteration of vehicle planning and control strategies. As a result, autonomous vehicles cannot make reasonable decisions in the face of untrained scenarios.

[0061] Cutting in from adjacent lanes to the front of the main vehicle is a common occurrence in road traffic. If autonomous vehicles cannot accurately predict or identify this behavior, and lack appropriate control strategies for such conflicts, rear-end collisions are highly likely. Furthermore, these cut-in scenarios often involve not only the two vehicles themselves but also other vehicles obscured by the roadside sensors. This is because vehicle-mounted cameras and sensors are susceptible to limitations imposed by weather conditions and visibility, resulting in limited detection range and long-tail issues. The applicant recognizes that roadside LiDAR can capture high-resolution 3D point cloud data of all participating vehicles and their surroundings, unaffected by weather or lighting conditions. Therefore, utilizing roadside LiDAR and other roadside sensors to create fused roadside perception data for cutting-in scenario selection not only reconstructs the cutting-in behavior of the vehicle and the main vehicle but also the complex states of all road users, facilitating the iteration of autonomous driving planning and control strategies.

[0062] Based on this, this application provides a method for building a scenario library for entry scenarios. When building the scenario library, conflicting vehicles and lane-changing vehicles are first screened based on vehicle behavior within the target road segment, resulting in sets of conflicting vehicles and lane-changing vehicles. Then, these sets are combined to screen for vehicles exhibiting entry scenarios. This identifies and filters vehicles that engage in lane-changing and cutting-in behaviors with dangerous conflict potential as vehicles with entry scenarios, improving the accuracy of entry scenario screening and thus enhancing the data quality of entry vehicle information. Consequently, the quality of the built scenario library is improved. Furthermore, the perception data is collected through roadside sensors, making the perception data more comprehensive. This results in a more comprehensive and complete scenario library built based on this comprehensive perception data, further enhancing the quality of the entry scenario library. Of course, the technical solutions provided in this application are not limited to solving the above problems and also have other technical effects, which can be seen in the following embodiments.

[0063] It should be noted that the beneficial effects or technical problems solved by the embodiments of this application are not limited to this one, but may also be other implicit or related problems. For details, please refer to the description of the embodiments below.

[0064] The following method for building a scene library for entry scenarios, provided in the embodiments of this application, can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or located in the cloud or on other network servers. Terminal 102 can be a roadside sensor, such as a camera, millimeter-wave radar, lidar, GPS (Global Positioning System) receiver, ultrasonic sensor, etc. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0065] The technical solution of this application and how it solves the above-mentioned technical problems will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0066] In one exemplary embodiment, such as Figure 2 As shown, a method for building a scene library by entering a scene is provided, and this method is applied to... Figure 1 Taking the server in the example, the explanation includes the following steps 201 to 203. Wherein:

[0067] S201, based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0068] The target road segment range refers to a specific road section, which can be set by the user according to actual needs. Roadside sensors are devices installed beside roads or on lampposts to sense the movement of vehicles and pedestrians, as well as road condition information. Roadside sensors include cameras, millimeter-wave radar, lidar, ultrasonic sensors, inertial measurement units (IMUs), and GPS receivers. Cameras capture image information on the road, identifying traffic signs, traffic lights, lane markings, etc., helping vehicles understand their surroundings. Millimeter-wave radar uses the millimeter-wave band for ranging and positioning, possessing the ability to penetrate fog, smoke, dust, and other harsh environments, thus performing excellently in adverse weather conditions. LiDAR acquires three-dimensional information of the surrounding environment by emitting laser light and receiving reflected signals, commonly used for high-precision environmental perception. Ultrasonic sensors detect the distance and relative velocity of objects by sending and receiving sound waves. An inertial measurement unit, composed of accelerometers, gyroscopes, and magnetometers, measures the motion state of objects in inertial space, including velocity, acceleration, and direction. A GPS receiver receives satellite positioning information, providing precise vehicle position and direction of travel.

[0069] The raw perception data consists of driving information collected by each vehicle from roadside sensors within the target road segment, including but not limited to vehicle positioning data, trajectory data, and possible vehicle energy consumption and performance data.

[0070] A conflict vehicle assembly refers to a group of vehicles on a road whose actions (such as changing direction, changing speed, or sudden stopping) could potentially cause a traffic accident. A lane-changing vehicle assembly refers to a group of vehicles that are changing lanes.

[0071] In this embodiment, raw perception data is acquired through multiple roadside sensors. Target detection and fusion tracking algorithms are used to obtain the timestamp, target ID, center point x-coordinate, center point y-coordinate, center point z-coordinate, length, width, height, speed, confidence level, heading angle, longitude, latitude, and frame number information for each vehicle at each time point. A coordinate system is established for the road, and the vehicle's latitude and longitude information is converted to obtain the lane number, lane area coordinates, lane centerline, and vehicle coordinates in the road coordinate system for each lane within the road segment. The road coordinate system coordinates {(x1,y1),(x2,y2),(x3,y3)...}, speed (v1, v2, v3...) time series information, and the converted road structure map information for each timetamp (t1, t2, t3...) are used as input data to filter vehicle behavior within the target road segment.

[0072] When filtering conflicting vehicles, pre-set conflicting vehicle filtering rules can be used. For example, conflicting vehicles can be filtered based on whether the distance between two vehicles is less than a preset threshold. Taking any two adjacent vehicles as an example, the distance between the two vehicles is calculated based on their road coordinate system coordinates. If the calculated distance is less than the preset threshold, the two adjacent vehicles are considered conflicting vehicles and added to the conflicting vehicle set.

[0073] When filtering lane-changing vehicles, the filtering can be based on whether each vehicle has engaged in lane-changing behavior. For example, for any given vehicle, the lane in the current frame and the lane in the previous frame are determined based on the vehicle's road coordinate system coordinates and lane coordinate area. If the lane in the current frame is different from the lane in the previous frame, the vehicle is identified as a lane-changing vehicle and added to the lane-changing vehicle set.

[0074] S202, based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut into the target road segment.

[0075] In this embodiment of the application, after filtering vehicle behavior within the target road segment to identify conflicting vehicles and lane-changing vehicles, vehicles with entry scenarios are selected from the conflicting vehicle set and the lane-changing vehicle set. For example, if a vehicle is the leading vehicle in a pair of conflicting vehicles, and this vehicle is in the lane-changing vehicle set and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then it is determined that there are entry vehicles. These two vehicles are identified as vehicles with entry scenarios, and the vehicle information of these two vehicles is recorded as entry vehicle information.

[0076] S203, build a scenario library for entry scenarios based on the information of the vehicle being entered.

[0077] After identifying vehicles that can be used in specific scenarios, a scenario library is built based on the vehicle information. For example, the scenario library is constructed based on the identified vehicle information, vehicle information of surrounding vehicles, and environmental data (such as weather data).

[0078] The scenario library construction method for entry scenarios provided in this application first filters vehicle behavior within the target road segment based on raw perception data collected by various roadside sensors within the target road segment, determining conflict vehicle sets and lane-changing vehicle sets. Then, based on the conflict vehicle sets and lane-changing vehicle sets, it determines entry vehicle information for entry scenarios within the target road segment. Finally, based on the entry vehicle information, it constructs a scenario library for entry scenarios. The raw perception data consists of driving information of multiple vehicles within the target road segment during their journey. In this method, when building the scenario library for entry scenarios, conflicting vehicles and lane-changing vehicles are first screened out based on vehicle behavior within the target road segment, resulting in sets of conflicting vehicles and lane-changing vehicles. Then, the sets of conflicting vehicles and lane-changing vehicles are combined to screen for entry vehicles that exist in entry scenarios. This identifies and filters vehicles that engage in lane-changing and cutting-in behaviors and exhibit dangerous conflict behaviors as entry vehicles with existing entry scenarios, improving the accuracy of entry scenario screening and thus improving the data quality of entry vehicle information. Consequently, the quality of the constructed scenario library for entry scenarios is enhanced. Furthermore, the perception data is collected through roadside sensors, making the perception data more comprehensive. This results in a more comprehensive and complete data library for entry scenarios built based on comprehensive perception data, further improving the quality of the entry scenario library.

[0079] Based on the above embodiments, an embodiment of the process for screening conflicting vehicle sets is provided for illustration.

[0080] In one exemplary embodiment, such as Figure 3 As shown, based on the raw perception data collected by various roadside sensors within the target road segment, vehicle behavior within the target road segment is filtered to determine the set of conflicting vehicles, including:

[0081] S301, based on the current and previous center point coordinates of each vehicle in the original perception data, filter multiple candidate conflict vehicle pairs.

[0082] In this embodiment of the application, the vehicle information collected consists of many frames, each frame having a timestamp. The vehicle information in each frame is compared with the vehicle information in the previous frame to filter out candidate conflicting vehicle pairs.

[0083] In one embodiment, the process of screening multiple candidate conflict vehicle pairs includes the following steps:

[0084] Step 1: For any two adjacent vehicles, determine the distance between their current center points based on their current center point coordinates; and determine the distance between their previous center points based on their previous center point coordinates.

[0085] Adjacent vehicles refer to two adjacent vehicles in the same motor vehicle lane.

[0086] For any two adjacent vehicles, obtain the coordinates of their center points at the current moment, and calculate the distance between their center points at the current moment using the distance formula and their coordinates. Similarly, obtain the coordinates of their center points at the previous moment, and calculate the distance between their center points at the previous moment using the distance formula and their coordinates.

[0087] Step 2: If the distance between the center points at the current time is less than the distance between the center points at the previous time, then the two adjacent vehicles are identified as a candidate conflict vehicle pair.

[0088] The distance between the center points of two adjacent vehicles at the previous moment is compared with the distance between their center points at the current moment. If the distance between the center points of the two adjacent vehicles at the current moment is less than the distance between their center points at the previous moment, it is determined that there is a possibility of conflict between the two vehicles, and the two adjacent vehicles are identified as a candidate conflict vehicle pair.

[0089] In this embodiment, the distance between the center points of two adjacent vehicles at the current time is calculated based on their center point coordinates at the current time, and the distance between their center points at the previous time is also calculated based on their center point coordinates at the previous time. Thus, based on the magnitude of the current center point distance and the previous center point distance between the two adjacent vehicles, candidate conflict vehicle pairs are screened, providing a data basis for the screening of conflict vehicles.

[0090] S302, determine the conflict time of each candidate vehicle pair based on the center point coordinates of each vehicle in the multiple candidate vehicle pairs at the current time.

[0091] For example, the conflict time of each candidate vehicle pair is determined based on the center point coordinates of each vehicle in the multiple candidate vehicle pairs at the current time, including:

[0092] Step 1: For any candidate conflict vehicle pair, obtain the coordinates of the intersection point of the trajectories of the two conflicting vehicles in the candidate conflict vehicle pair.

[0093] For any conflicting vehicle, use that conflicting vehicle Previous moment center point coordinates and the current moment center point coordinates By fitting the linear equation, the vehicle coordinates in the road coordinate system are obtained. The real-time trajectory slope-intercept form of the linear equation expression After obtaining the trajectory equations of the two conflicting vehicles using the above method, the two trajectory equations are combined to form a system of equations. Solving this system of equations yields the coordinates of the intersection point of the trajectories of the two conflicting vehicles. .

[0094] Step 2: Based on the coordinates of the intersection point of the trajectories of the two conflicting vehicles and the coordinates of the center points of the two conflicting vehicles, determine the distance between the coordinates of the center points of the two conflicting vehicles and the coordinates of the intersection point of their trajectories.

[0095] For any conflicting vehicle, according to the distance formula, substitute the coordinates of the center point of the conflicting vehicle and the coordinates of the intersection of the trajectory into the distance formula to obtain the distance between the coordinates of the center point of the conflicting vehicle and the coordinates of the intersection of the trajectory.

[0096] Step 3: Determine the conflict time of the candidate vehicle pair based on the distance between the center point coordinates and the trajectory intersection coordinates of the two conflicting vehicles, and the speed of the two conflicting vehicles at the current moment.

[0097] Among them, conflict time is a vehicle conflict index, which refers to the time required to travel from the location where the conflict begins to the conflict area at the speed at the moment the conflict occurs.

[0098] In this embodiment of the application, the conflict time of the candidate conflict vehicle pair can be calculated using the following formula (1).

[0099] (1)

[0100] in, For conflict time; and They are Time vehicle and The distance between the center point coordinates and the intersection of the trajectory. and yes Time vehicle and The speed.

[0101] The conflict times calculated using data from two conflicting vehicles may not be exactly the same. You can take the same number of decimal places as the final conflict time.

[0102] In this embodiment, an optional method is provided to quickly determine the conflict time of candidate conflict vehicle pairs. By calculating the conflict time of each candidate conflict vehicle pair, a screening basis is provided for the selection of conflict vehicles.

[0103] S303, determine the set of conflicting vehicles based on the conflict time of each candidate conflicting vehicle pair.

[0104] Once the conflict time of each candidate vehicle pair is determined, conflict vehicles can be selected from multiple candidate vehicle pairs based on the conflict time to construct a set of conflict vehicles.

[0105] For example, the distance between the center points of each candidate conflicting vehicle pair at the current time is obtained; candidate conflicting vehicle pairs whose current center point distance is less than a preset distance threshold and whose conflict time is less than a preset time threshold are identified; and a set of conflicting vehicles is determined based on the candidate conflicting vehicle pairs whose current center point distance is less than the preset distance threshold and whose conflict time is less than the preset time threshold.

[0106] The distance between the center points of each candidate conflicting vehicle pair at the current moment is calculated based on the coordinates of the center points of the two conflicting vehicles in each candidate conflicting vehicle pair at the current moment, as described in the previous embodiment, and will not be repeated here. Further, for any candidate conflicting vehicle pair, the distance between the center points of the candidate conflicting vehicle pair at the current moment is compared with a preset distance threshold, and the conflict time of the candidate conflicting vehicle pair is compared with a preset time threshold. If the distance between the center points of the candidate conflicting vehicle pair at the current moment is less than the preset distance threshold and the conflict time is less than the preset time threshold, then the candidate conflicting vehicle pair is determined to be conflicting vehicles, and the IDs of the two conflicting vehicles, the conflict time, and the timestamp of the conflict occurrence are recorded and added to the conflicting vehicle set.

[0107] The scenario library construction method for entry scenarios provided in this application first filters multiple candidate conflict vehicle pairs based on the current and previous center point coordinates of each vehicle in the original perception data. Then, based on the current center point coordinates of each conflicting vehicle in the multiple candidate conflict vehicle pairs, the conflict time of each candidate conflict vehicle pair is determined. Finally, based on the conflict time of each candidate conflict vehicle pair, the conflict vehicle set is determined. In this method, candidate conflict vehicle pairs are first filtered from each vehicle in the target road segment range, and then vehicle conflict index (conflict time) is calculated to filter the final conflicting vehicles from multiple candidate conflict vehicle pairs based on the conflict time, thereby constructing a conflict vehicle set, which improves the efficiency and accuracy of conflict vehicle filtering.

[0108] Based on the above embodiments, an embodiment of the process for determining the set of lane-changing vehicles is provided for illustration.

[0109] In one exemplary embodiment, such as Figure 4 As shown, based on the raw perception data collected by various roadside sensors within the target road segment, vehicle behavior within the target road segment is filtered to determine the set of lane-changing vehicles, including:

[0110] S401, based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map, determine the lane number information of each vehicle in each frame.

[0111] The coordinate range of the lane map includes the coordinate range of each lane.

[0112] In this embodiment, for any vehicle in any frame, the coordinates of the vehicle's center point are compared with the coordinate range of each lane in the lane map to obtain the lane number information of the vehicle in the current frame. The lane in which the vehicle's center point coordinates fall determines the lane in which the vehicle is located.

[0113] S402, determine the set of vehicles changing lanes based on the lane number information of each vehicle in each frame.

[0114] The lane number information of each vehicle in the current frame is compared with the lane change information of the previous frame to filter out the lane-changing vehicles.

[0115] For example, based on the lane number information of each vehicle in each frame, candidate vehicles whose lane number information in the current frame is different from that in the previous frame are obtained; based on the candidate vehicles whose lane number information in the current frame is different from that in the previous frame, the set of lane-changing vehicles is determined.

[0116] For any vehicle in any frame, obtain the lane number information of the vehicle in the current frame and the lane-changing vehicle information of the vehicle in the previous frame. Then compare the lane number information of the vehicle in the current frame with the lane-changing vehicle information of the previous frame. If the lane number information of the vehicle in the current frame is different from the lane-changing vehicle information of the previous frame, then the vehicle is identified as a lane-changing vehicle, and the vehicle's ID and the timestamp of the current frame are recorded in the lane-changing vehicle set.

[0117] The scenario library construction method for entry scenarios provided in this application first determines the lane number information of each vehicle in each frame based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map. Then, based on the lane number information of each vehicle in each frame, a set of lane-changing vehicles is determined. In this method, the lane number information of each vehicle in each frame is determined by the center point coordinates and lane coordinate range of each vehicle, which serves as the data basis for filtering lane-changing vehicles. Then, based on the lane number information of each vehicle in each frame, vehicles whose lanes in the current frame are different from those in the previous frame are filtered out, and a set of lane-changing vehicles is constructed based on these lane-changing vehicles, which improves the filtering efficiency of lane-changing vehicles and lays the foundation for entry scenario filtering.

[0118] Based on the above embodiments, an embodiment is provided to illustrate the process of determining the information of vehicles entering a cutting-in scenario within the target road segment range.

[0119] In one exemplary embodiment, such as Figure 5 As shown, based on the conflict vehicle set and the lane-changing vehicle set, the information of vehicles cutting into the target road segment is determined, including:

[0120] S501, For any pair of conflicting vehicles in the set of conflicting vehicles, determine the front conflicting vehicle and the rear conflicting vehicle in the pair of conflicting vehicles based on the coordinates of the center points of the two conflicting vehicles in the pair of conflicting vehicles.

[0121] For any pair of conflicting vehicles A and B, calculate the vector based on the coordinates of the center points of the two conflicting vehicles. And then calculate If the cosine of the vector relative to the direction of travel of the lane is less than 0, then car A is in front; otherwise, car B is in front.

[0122] S502, if the vehicle in the preceding conflict is within the set of vehicles changing lanes and the difference between the time of the lane change of the vehicle in the preceding conflict and the time of the conflict is less than a preset threshold, then it is determined that the conflicting vehicle pair has an entry scenario.

[0123] After identifying the vehicles involved in the preceding and following conflicts, it is determined whether the vehicle involved in the preceding conflict has a lane-changing record. If the vehicle involved in the preceding conflict is within the set of vehicles that changed lanes and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then the vehicles involved in the conflict have an entry scenario. The lane-changing time is obtained from the set of vehicles that changed lanes, and the time of the conflict is obtained from the set of vehicles that experienced the conflict.

[0124] S503 identifies the lane-changing time of the vehicle in the preceding conflict, the vehicle identification of the vehicle in the preceding conflict, and the vehicle identification of the vehicle in the following conflict as the entry vehicle information.

[0125] After identifying the conflicting vehicle pairs in the cutting-in scenario, the lane change time of the preceding conflicting vehicle is used as the timestamp of the cutting-in action. The timestamp of the cutting-in action, the vehicle identifier of the preceding conflicting vehicle (the main vehicle being cut in), and the vehicle identifier of the following conflicting vehicle (the vehicle that performed the cutting-in action) are used as the cutting-in vehicle information, and output as a single CSV file to obtain the cutting-in scenario dataset.

[0126] In the scenario library construction method for entry scenarios provided in this application embodiment, for any pair of conflicting vehicles in the conflicting vehicle set, the preceding and following conflicting vehicles are determined based on the center point coordinates of the two conflicting vehicles in the pair. If the preceding conflicting vehicle is within the lane-changing vehicle set and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then the conflicting vehicle pair is determined to have an entry scenario. Then, the lane-changing time of the preceding conflicting vehicle, the vehicle identifier of the preceding conflicting vehicle, and the vehicle identifier of the following conflicting vehicle are determined as entry vehicle information. In this method, the preceding and following vehicles are first identified from the conflicting vehicles. Then, if the preceding vehicle has a lane-changing behavior record and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then an entry scenario is determined to exist. The vehicle information of vehicles with entry scenarios is recorded to quickly obtain entry vehicle information, improving the accuracy of entry scenario selection and providing data support for improving the safety and reliability of autonomous driving.

[0127] Based on the above embodiments, an embodiment is provided to illustrate the process of building a scenario library for entry scenarios.

[0128] In one exemplary embodiment, such as Figure 6 As shown, based on the vehicle information being entered, a scenario library for entry scenarios is built, including:

[0129] S601, based on the vehicle information, obtain the scene start timestamp and scene end timestamp.

[0130] In this embodiment of the application, the scene start timestamp and the scene end timestamp are the same time, which can be obtained from the vehicle information.

[0131] S602, based on the scene start timestamp and scene end timestamp, obtain the target trajectory data and target weather data corresponding to the vehicle entering the scene.

[0132] Among them, the target trajectory data and target weather data corresponding to each vehicle entering the scene refer to all vehicle trajectory data and weather data from the preset time before the scene begins to the preset time after the scene ends.

[0133] The preset time is a pre-set time value required for the scene, which can be set according to actual needs. For example, it can be set to 30 seconds.

[0134] Roadside sensors collect trajectory data (fused tracking data) from all vehicles and weather data, storing them in a preset storage structure to obtain a trajectory dataset and a weather dataset. The trajectory dataset contains trajectory data for all targets on the road within the time interval [scene start timestamp - required time value, scene start timestamp + required time value], serving as the target trajectory data. Similarly, the weather dataset contains weather data within the minute interval [scene start timestamp - required time value, scene start timestamp + required time value], serving as the target weather data.

[0135] S603 builds a scenario library based on target trajectory data, target weather data, and information on vehicles entering the area.

[0136] After determining the target trajectory data and target weather data, scene slice data is generated by combining the previously stored vehicle information. This data is then stored in a database to construct a scene library for the scene being cut off. The data includes information such as [main target ID, whether it is the main target of the scene, timestamp, center point x-coordinate, center point y-coordinate, center point z-coordinate, length, width, height, speed, confidence level, heading angle, longitude, latitude, and frame number]. For each frame, each vehicle generates such a dataset. All data for each frame is stored in a CSV file named "Entry Scene Type Value_Scene Number_Frame Number". All slice data for a complete scene are stored in a folder named "Entry Scene Type Value_Scene Number". The entry scene type value is always the same, the scene number consists of the generated ID plus the specific date, and the frame number is obtained by numbering each frame.

[0137] The method for building a scenario library for entry scenarios provided in this application embodiment obtains the scenario start timestamp and scenario end timestamp based on the entry vehicle information. Then, based on the scenario start timestamp and scenario end timestamp, it obtains the target trajectory data and target weather data corresponding to the entry vehicles in the entry scenarios. The target trajectory data and target weather data corresponding to each entry vehicle refer to all vehicle trajectory data and weather data from a preset time before the start of the entry scenario to a preset time after the end of the entry scenario. A scenario library is built based on the target trajectory data, target weather data, and entry vehicle information. This method provides an optional way to quickly build a scenario library for entry scenarios. By obtaining all vehicle trajectory data and weather data corresponding to the entry vehicles and combining them with the entry vehicle information, a more comprehensive and complete entry scenario dataset is formed, thereby building a scenario library for entry scenarios, improving the data quality of the scenario library, and providing a reliable data foundation for autonomous driving.

[0138] In addition, in an exemplary embodiment, this application also provides a method for building a scene library for entry into scenarios, such as... Figure 7 As shown, it may include the following steps:

[0139] S701, based on the current and previous center point coordinates of each vehicle in the original perception data, filters out multiple candidate conflict vehicle pairs.

[0140] Optionally, for any two adjacent vehicles, the distance between their current center points is determined based on their current center point coordinates; and the distance between their previous center points is determined based on their previous center point coordinates; if the current center point distance is less than the previous center point distance, then the two adjacent vehicles are identified as a candidate conflict vehicle pair.

[0141] S702, determine the conflict time of each candidate vehicle pair based on the center point coordinates of each vehicle in the current time.

[0142] Optionally, for any candidate conflict vehicle pair, obtain the coordinates of the trajectory intersection point of the two conflict vehicles in the candidate conflict vehicle pair; determine the distance between the coordinates of the center point of the two conflict vehicles and the coordinates of the trajectory intersection point based on the coordinates of the trajectory intersection point of the two conflict vehicles and the coordinates of the center point of the two conflict vehicles; determine the conflict time of the candidate conflict vehicle pair based on the distance between the coordinates of the center point of the two conflict vehicles and the coordinates of the trajectory intersection point, and the speed of the two conflict vehicles at the current time.

[0143] S703, determine the set of conflicting vehicles based on the conflict time of each candidate conflicting vehicle pair.

[0144] Optionally, obtain the current distance between the center points of each candidate conflicting vehicle pair; determine the candidate conflicting vehicle pairs whose current distance between the center points is less than a preset distance threshold and whose conflict time is less than a preset time threshold; and determine the conflicting vehicle set based on the candidate conflicting vehicle pairs whose current distance between the center points is less than the preset distance threshold and whose conflict time is less than the preset time threshold.

[0145] S704 determines the lane number information of each vehicle in each frame based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map.

[0146] S705 determines the set of vehicles changing lanes based on the lane number information of each vehicle in each frame.

[0147] Optionally, based on the lane number information of each vehicle in each frame, candidate vehicles whose lane number information in the current frame is different from that in the previous frame are obtained; based on the candidate vehicles whose lane number information in the current frame is different from that in the previous frame, the set of lane-changing vehicles is determined.

[0148] S706, based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut into the target road segment.

[0149] Optionally, for any pair of conflicting vehicles in the conflicting vehicle set, the preceding and following conflicting vehicles are determined based on the center point coordinates of the two conflicting vehicles in the pair. If the preceding conflicting vehicle is within the lane-changing vehicle set and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then it is determined that the conflicting vehicle pair has an entry scenario. The lane-changing time of the preceding conflicting vehicle, the vehicle identifier of the preceding conflicting vehicle, and the vehicle identifier of the following conflicting vehicle are determined as entry vehicle information.

[0150] S707 obtains the scene start timestamp and scene end timestamp based on the vehicle information.

[0151] S708 obtains the target trajectory data and target weather data corresponding to the vehicles entering the scene based on the scene start timestamp and scene end timestamp.

[0152] Among them, the target trajectory data and target weather data corresponding to each vehicle entering the scene refer to all vehicle trajectory data and weather data from the preset time before the scene begins to the preset time after the scene ends.

[0153] S709 builds a scenario library based on target trajectory data, target weather data, and information on vehicles entering the area.

[0154] The processes S701-S709 described above can be found in the description of the above method embodiments, and their implementation principles and technical effects are similar, so they will not be repeated here.

[0155] In addition, in one embodiment, the process of selecting entry scenarios and building a scenario library is described.

[0156] like Figure 8 As shown, lane information / map information is used as input data for entry scene filtering. When filtering entry scenes, conflict target pairs and conflict time sets, as well as lane change targets and lane change time sets are first determined. Then, based on the conflict target pairs and lane change targets, targets with entry scenes are determined and the entry scene filtering results are output. After that, the entry scene library is built by combining weather data and multi-sensor fusion tracking data.

[0157] In this embodiment, when building the scenario library for entry scenarios, conflicting vehicles and lane-changing vehicles are first screened out based on vehicle behavior within the target road segment, resulting in sets of conflicting vehicles and lane-changing vehicles. Then, the sets of conflicting vehicles and lane-changing vehicles are combined to screen for entry vehicles with entry scenarios. This identifies and filters vehicles that engage in lane-changing and cutting-in behaviors and exhibit dangerous conflict behaviors as entry vehicles with entry scenarios, improving the accuracy of entry scenario screening and thus improving the data quality of entry vehicle information. Consequently, the quality of the built scenario library for entry scenarios is improved. Furthermore, the perception data is collected through roadside sensors, making the perception data more comprehensive. This results in a more comprehensive and complete data library for entry scenarios built based on comprehensive perception data, further enhancing the quality of the entry scenario library.

[0158] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0159] Based on the same inventive concept, this application also provides a scene library building device for implementing the scene library building method for the aforementioned scene entry scenario. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more scene library building device embodiments provided below can be found in the limitations of the scene library building method for the aforementioned scene entry scenario, and will not be repeated here.

[0160] In one exemplary embodiment, such as Figure 9 As shown, a scene library building device for entry scenarios is provided, including: a vehicle filtering module 10, an entry scenario filtering module 20, and a scene library building module 30, wherein:

[0161] The vehicle screening module 10 is used to screen vehicle behavior within the target road segment based on raw perception data collected by various roadside sensors within the target road segment, and to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0162] The entry scene filtering module 20 is used to determine the entry vehicle information of the target road segment range that has an entry scene based on the conflict vehicle set and the lane-changing vehicle set.

[0163] The scenario library building module 30 is used to build a scenario library for the entry scenario based on the entry vehicle information.

[0164] In one embodiment, the vehicle screening module 10 is further configured to:

[0165] Based on the current and previous center point coordinates of each vehicle in the original sensing data, multiple candidate conflict vehicle pairs are selected; based on the current center point coordinates of each conflict vehicle in the multiple candidate conflict vehicle pairs, the conflict time of each candidate conflict vehicle pair is determined; based on the conflict time of each candidate conflict vehicle pair, the conflict vehicle set is determined.

[0166] In one embodiment, the vehicle screening module 10 is further configured to:

[0167] For any two adjacent vehicles, determine the distance between their current center points based on their current center point coordinates; and determine the distance between their previous center points based on their previous center point coordinates. If the current center point distance is less than the previous center point distance, then the two adjacent vehicles are determined to be a candidate conflict vehicle pair.

[0168] In one embodiment, the vehicle screening module 10 is further configured to:

[0169] For any candidate vehicle pair, obtain the coordinates of the intersection point of the trajectories of the two vehicles in the candidate vehicle pair; determine the distance between the coordinates of the center point of the two vehicles and the coordinates of the intersection point based on the coordinates of the intersection point of the trajectories of the two vehicles; determine the conflict time of the candidate vehicle pair based on the distance between the center point of the two vehicles and the coordinates of the intersection point of the trajectories, and the speed of the two vehicles at the current moment.

[0170] In one embodiment, the vehicle screening module 10 is further configured to:

[0171] Obtain the current distance between the center points of each candidate conflicting vehicle pair; identify candidate conflicting vehicle pairs whose current distance between the center points is less than a preset distance threshold and whose conflict time is less than a preset time threshold; determine the conflicting vehicle set based on the candidate conflicting vehicle pairs whose current distance between the center points is less than the preset distance threshold and whose conflict time is less than the preset time threshold.

[0172] In one embodiment, the vehicle screening module 10 is further configured to:

[0173] Based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map, the lane number information of each vehicle in each frame is determined; based on the lane number information of each vehicle in each frame, the set of vehicles changing lanes is determined.

[0174] In one embodiment, the vehicle screening module 10 is further configured to:

[0175] Based on the lane number information of each vehicle in each frame, obtain candidate vehicles whose lane number information in the current frame is different from that in the previous frame; based on the candidate vehicles whose lane number information in the current frame is different from that in the previous frame, determine the set of lane-changing vehicles.

[0176] In one embodiment, the above-mentioned scene selection module 20 is further configured to:

[0177] For any pair of conflicting vehicles in the conflicting vehicle set, the preceding and following conflicting vehicles are determined based on the center point coordinates of the two conflicting vehicles in the pair. If the preceding conflicting vehicle is within the lane-changing vehicle set and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then the conflicting vehicle pair is determined to have an entry scenario. The lane-changing time of the preceding conflicting vehicle, the vehicle identifier of the preceding conflicting vehicle, and the vehicle identifier of the following conflicting vehicle are determined as the entry vehicle information.

[0178] In one embodiment, the scene library building module 30 is further used for:

[0179] Based on the information of the vehicle entering the scene, obtain the scene start timestamp and scene end timestamp; based on the scene start timestamp and scene end timestamp, obtain the target trajectory data and target weather data corresponding to the vehicle entering the scene; the target trajectory data and target weather data corresponding to each vehicle entering the scene refer to all vehicle trajectory data and weather data from the preset time before the start of the scene entering the scene to the preset time after the end of the scene entering the scene; build a scene library based on the target trajectory data, target weather data and vehicle entering the scene information.

[0180] Each module in the aforementioned scene library building device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0181] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores scene library construction data for entry scenarios. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a scene library construction method for entry scenarios.

[0182] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0183] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0184] Based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0185] Based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut in within the target road segment;

[0186] Based on the vehicle information, a scenario library for entry scenarios is built.

[0187] The implementation principles and technical effects of each step in the processor implementation of this application embodiment are similar to those of the above-mentioned scene library construction method for entering the scene, and will not be repeated here.

[0188] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0189] Based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0190] Based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut in within the target road segment;

[0191] Based on the vehicle information, a scenario library for entry scenarios is built.

[0192] The implementation principles and technical effects of each step in the computer program executed by the processor in this embodiment are similar to those of the above-mentioned method for building a scene library for entering a scene, and will not be repeated here.

[0193] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0194] Based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey.

[0195] Based on the conflict vehicle set and the lane-changing vehicle set, determine the information of vehicles that have cut in within the target road segment;

[0196] Based on the vehicle information, a scenario library for entry scenarios is built.

[0197] The implementation principles and technical effects of each step in the computer program executed by the processor in this embodiment are similar to those of the above-mentioned method for building a scene library for entering a scene, and will not be repeated here.

[0198] It should be noted that the personnel information (including but not limited to personnel and equipment information, personnel personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the personnel or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0199] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0200] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0201] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for building a scene library for entering a scene, characterized in that, The method includes: Based on the raw perception data collected by various roadside sensors within the target road segment, the vehicle behavior within the target road segment is filtered to determine the conflicting vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey. Based on the conflict vehicle set and the lane-changing vehicle set, determine the cutting vehicle information within the target road segment range where there is a cutting scenario; Based on the vehicle information, a scenario library for entry scenarios is built.

2. The method according to claim 1, characterized in that, The process of filtering vehicle behavior within the target road segment based on raw perception data collected by multiple roadside sensors within the target road segment to determine the conflicting vehicle set includes: Based on the current and previous center point coordinates of each vehicle in the original sensing data, multiple candidate conflict vehicle pairs are filtered. The conflict time of each candidate vehicle pair is determined based on the center point coordinates of each vehicle in the multiple candidate vehicle pairs at the current time. The set of conflicting vehicles is determined based on the conflict time of each of the candidate conflicting vehicle pairs.

3. The method according to claim 2, characterized in that, The step of filtering multiple candidate conflict vehicle pairs based on the current and previous center point coordinates of each vehicle in the original sensing data includes: For any two adjacent vehicles, the distance between their current center points is determined based on their current center point coordinates; and the distance between their previous center points is determined based on their previous center point coordinates. If the distance between the center points at the current time is less than the distance between the center points at the previous time, then the two adjacent vehicles are determined to be a candidate conflict vehicle pair.

4. The method according to claim 2, characterized in that, The step of determining the conflict time of each candidate vehicle pair based on the center point coordinates of each vehicle in the multiple candidate vehicle pairs at the current time includes: For any candidate conflict vehicle pair, obtain the coordinates of the intersection point of the trajectories of the two conflict vehicles in the candidate conflict vehicle pair; Based on the coordinates of the intersection point of the trajectories of the two conflicting vehicles and the coordinates of the center points of the two conflicting vehicles, determine the distance between the coordinates of the center points of the two conflicting vehicles and the coordinates of the intersection point of their trajectories; The conflict time of the candidate vehicle pair is determined based on the distance between the center point coordinates of the two conflicting vehicles and the coordinates of the intersection of their trajectories, as well as the speeds of the two conflicting vehicles at the current moment.

5. The method according to claim 2, characterized in that, Determining the set of conflicting vehicles based on the conflict time of each of the candidate conflicting vehicle pairs includes: Obtain the current distance between the center points of each candidate conflict vehicle pair; Determine candidate conflict vehicle pairs whose current center point distance is less than a preset distance threshold and whose conflict time is less than a preset time threshold. The set of conflicting vehicles is determined based on candidate conflicting vehicle pairs whose current center point distance is less than a preset distance threshold and whose conflict time is less than a preset time threshold.

6. The method according to any one of claims 1-5, characterized in that, The process involves filtering vehicle behavior within the target road segment based on raw perception data collected from multiple roadside sensors to determine a set of lane-changing vehicles, including: Based on the center point coordinates of each vehicle in the original perception data and the coordinate area range of the lane map, the lane number information of each vehicle in each frame is determined. The set of vehicles changing lanes is determined based on the lane number information of each vehicle in each frame.

7. The method according to claim 6, characterized in that, Determining the set of lane-changing vehicles based on the lane number information of each vehicle in each frame includes: Based on the lane number information of each vehicle in each frame, obtain candidate vehicles whose lane number information in the current frame is different from that in the previous frame. The set of lane-changing vehicles is determined based on the candidate vehicles whose lane number information in the current frame is different from that in the previous frame.

8. The method according to any one of claims 1-5, characterized in that, The step of determining the cutting-in vehicle information within the target road segment range based on the conflicting vehicle set and the lane-changing vehicle set includes: For any pair of conflicting vehicles in the set of conflicting vehicles, the front conflicting vehicle and the rear conflicting vehicle are determined based on the center point coordinates of the two conflicting vehicles in the pair. If the vehicle in the previous conflict is within the set of lane-changing vehicles and the difference between the lane-changing time and the time of the conflict is less than a preset threshold, then it is determined that the conflicting vehicle pair has an entry scenario. The lane-changing time of the vehicle that was in the preceding conflict, the vehicle identification of the vehicle that was in the preceding conflict, and the vehicle identification of the vehicle that was in the following conflict are determined as the information of the vehicle that cut in.

9. The method according to any one of claims 1-5, characterized in that, The step of building a scenario library for entry scenarios based on the entry vehicle information includes: Based on the vehicle information, obtain the scene start timestamp and scene end timestamp; Based on the start timestamp and end timestamp of the scene, obtain the target trajectory data and target weather data corresponding to the vehicles entering the scene; the target trajectory data and target weather data corresponding to each vehicle entering the scene refer to all vehicle trajectory data and weather data from a preset time before the start of the scene to a preset time after the end of the scene. The scenario library is built based on the target trajectory data, the target weather data, and the information of the vehicle that entered the scene.

10. A scene library building device for entering a scene, characterized in that, The device includes: The vehicle filtering module is used to filter vehicle behavior within the target road segment based on raw perception data collected by various roadside sensors within the target road segment, and to determine the conflict vehicle set and the lane-changing vehicle set; the raw perception data is the driving information of multiple vehicles within the target road segment during their journey. The entry scenario filtering module is used to determine the entry vehicle information of the target road segment range that has an entry scenario based on the conflict vehicle set and the lane-changing vehicle set. The scenario library building module is used to build a scenario library for the entry scenario based on the entry vehicle information.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9.

13. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9.

Images