Vehicle intersection decision method, device, equipment and storage medium

Abstract

The present disclosure provides a vehicle intersection decision method, device, equipment and storage medium. The method comprises: in response to the acquired environmental data, determining that the autonomous vehicle enters the intersection and no traffic light signal is detected, determining the target vehicle and the state information of the target vehicle that need to be analyzed from the environmental data based on the set filtering rule, then determining the expected meeting state of the autonomous vehicle and the target vehicle based on the positional relationship between the first driving track of the target vehicle and the second driving track of the autonomous vehicle, and the shape parameter of the set part of the target vehicle, and finally determining the driving strategy of the autonomous vehicle based on the expected meeting state. The present application solves the problem of lack of stable and reliable decision method when the autonomous vehicle passes through the intersection without traffic lights, and ensures the driving efficiency and safety of the autonomous vehicle passing through the intersection.

Description

Technical Field

[0001] This disclosure relates to the field of autonomous driving technology, and in particular to a vehicle intersection decision-making method, apparatus, device, and storage medium. Background Technology

[0002] With the emergence of diverse transportation demands, autonomous driving technology is gradually being applied more widely. A significant challenge in the application of autonomous driving technology lies in driving at intersections. Whether in urban road environments or enclosed parks, intersections are essential passageways for manned and autonomous vehicles, pedestrians, and other vehicles merging, turning, and evacuating. Numerous factors influence vehicle safety, making decision-making difficult and easily leading to safety hazards. This is especially true for intersections without traffic lights, where the environment is even more complex, posing a significant challenge to the safe and efficient passage of autonomous vehicles.

[0003] In existing technologies, the decision-making of autonomous vehicles when driving at intersections is mainly based on the traffic lights. At intersections without traffic lights, the decision-making is based on learning methods. However, the results of these methods are not stable enough and cannot guarantee the safety of autonomous vehicles. Summary of the Invention

[0004] This disclosure provides a vehicle intersection decision-making method, apparatus, device, and storage medium to address the problem in the prior art of lacking a stable and reliable decision-making method when autonomous vehicles pass through intersections without traffic lights.

[0005] Firstly, this disclosure provides a vehicle intersection decision-making method, which includes:

[0006] Based on the acquired environmental data, it is determined that the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0007] Based on the set filtering rules, the target vehicle and its status information to be analyzed are determined from the environmental data. The status information includes the target vehicle's first driving trajectory and the shape parameters of the set parts.

[0008] Based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the designated parts of the target vehicle, the expected meeting state between the autonomous vehicle and the target vehicle is determined.

[0009] Based on the expected oncoming traffic situation, determine the driving strategy for autonomous vehicles.

[0010] As can be seen, by responding to the acquired environmental data, it is determined that the autonomous vehicle has entered an intersection without detecting any traffic light signals. Based on pre-defined filtering rules, the target vehicle and its state information are identified from the environmental data. Then, based on the target vehicle's first driving trajectory, the shape parameters of a designated location, and the positional relationship between the autonomous vehicle's second driving trajectory and the target vehicle, the expected meeting state between the autonomous vehicle and the target vehicle is determined. Finally, based on the expected meeting state, the autonomous vehicle's driving strategy is determined. Therefore, even at intersections without traffic lights, target vehicles affecting the autonomous vehicle's driving can be effectively identified. Based on the target vehicle's driving state and structural characteristics, it can be determined whether the autonomous vehicle needs to slow down or evade. This allows the autonomous vehicle to have reliable rules to ensure driving safety at intersections without traffic lights. By filtering target vehicles and determining driving strategies accordingly, the system ensures vehicle safety while minimizing the impact of environmental factors, thus guaranteeing the efficiency and safety of the autonomous vehicle's passage through intersections.

[0011] Optionally, based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the designated part of the target vehicle, the expected meeting state between the autonomous vehicle and the target vehicle is determined, including: based on the type of the target vehicle, determining the target structure with the largest width of the target vehicle as the designated part; using a point on the first driving trajectory as the center, generating a trajectory expansion circle corresponding to the first driving trajectory based on the width of the target structure and the designated expansion width, wherein the designated expansion width is determined based on the type of the target vehicle; and determining the expected meeting state based on the intersection of the second driving trajectory and the trajectory expansion circle.

[0012] As can be seen, by combining the trajectory expansion circle of the first driving trajectory with the intersection of the second driving trajectory to determine the expected meeting state, the possible state when the autonomous vehicle meets the target vehicle can be accurately determined graphically. The trajectory expansion circle is determined by multiple parameters to fully consider the structural characteristics of the target vehicle, thereby ensuring the accuracy of the determination of the possible state when meeting, and thus enabling the accurate determination of the driving strategy to ensure the safety of the autonomous vehicle.

[0013] Optionally, the expected meeting state is determined based on the intersection of the second driving trajectory and the trajectory expansion circle, including: determining the intersection point of the second driving trajectory and the first driving trajectory; determining the tangent point of the trajectory expansion circle and the second driving trajectory; determining the point closest to the autonomous vehicle among the intersection point and the tangent point as the conflict point; and determining the expected meeting state based on the target vehicle's action state, the autonomous vehicle's action state, and the conflict point.

[0014] It is evident that by identifying conflict points from multiple intersection and tangency points, and determining the expected meeting state based on the conflict points, the positions of the autonomous vehicle and the target vehicle in the earliest expected meeting state can be determined, thereby effectively avoiding the most urgent dangers and ensuring the safety of autonomous vehicle operation.

[0015] Optionally, the target vehicle's movement state includes a first speed, and the autonomous vehicle's movement state includes a second speed. Based on the target vehicle's movement state, the autonomous vehicle's movement state, and the conflict point, the expected meeting state is determined, including: determining the first moment when the head of the autonomous vehicle reaches the conflict point based on the second speed; determining the second moment when the head of the target vehicle reaches the conflict point and the third moment when the rear of the target vehicle reaches the conflict point based on the first speed; if the difference between the second moment and the first moment is greater than a first time threshold, the expected meeting state is determined to be that the autonomous vehicle passes the conflict point first and will not collide with the target vehicle; or, if the difference between the first moment and the third moment is greater than the second time threshold, the expected meeting state is determined to be that the target vehicle passes the conflict point first and will not collide with the autonomous vehicle; or, if the difference between the first moment and the second moment is less than the first time threshold, the expected passing state is determined to be that the target vehicle and the autonomous vehicle will collide.

[0016] It is evident that by accurately determining the expected oncoming traffic conditions based on different target vehicles and the driving status of autonomous vehicles, potential collisions can be avoided in advance, thus ensuring the safety of autonomous vehicles.

[0017] Optionally, based on the expected meeting state, the driving strategy of the autonomous vehicle is determined, including: if the expected meeting state is that the autonomous vehicle and the target vehicle will not collide, the driving strategy is determined to maintain the current driving state; if the expected meeting state is that the target vehicle and the autonomous vehicle will collide, the driving strategy of the autonomous vehicle is determined based on the first speed and the first driving trajectory of the target vehicle.

[0018] It is evident that determining the driving strategy based on whether a collision will occur between the target vehicle and the autonomous vehicle, maintaining the driving state when there is no risk of collision, avoiding ineffective adjustments, and improving the driving efficiency and stability of autonomous vehicles.

[0019] Optionally, if the expected oncoming situation is that a collision will occur between the target vehicle and the autonomous vehicle, the driving strategy of the autonomous vehicle is determined based on the first speed and the first driving trajectory of the target vehicle, including: if the first speed is higher than a set first speed threshold, the driving speed is adjusted based on a set deceleration parameter until the target vehicle passes the conflict point first and there is no collision with the autonomous vehicle; if the first speed is lower than the set first speed threshold, a safe meeting time is determined based on the distance between the target vehicle and the conflict point and the first speed; and the driving speed is re-determined based on the location of the autonomous vehicle and the safe meeting time.

[0020] It is evident that different driving strategies can effectively avoid collision risks in different scenarios and fully ensure the driving safety of autonomous vehicles.

[0021] Optionally, based on the set filtering rules, the target vehicle to be analyzed and its state information are determined from the environmental data, including: filtering vehicles whose distance from the autonomous vehicle is greater than a first set distance threshold; filtering vehicles identified in the environmental data that are located behind the autonomous vehicle based on the autonomous vehicle's driving direction; if the autonomous vehicle's driving direction is straight, filtering vehicles identified in the environmental data that are traveling in the same direction as the autonomous vehicle; or, if the autonomous vehicle's driving direction is left-turning, filtering vehicles identified in the environmental data that are traveling straight or turning right to the right of the autonomous vehicle; or, if the autonomous vehicle's driving direction is right-turning, filtering vehicles identified in the environmental data that are traveling straight or turning left to the left of the autonomous vehicle; identifying vehicles among the filtered vehicles whose angle with the autonomous vehicle's driving direction is within a set angle range as the target vehicles to be analyzed; and obtaining the state information of the target vehicles based on the environmental data.

[0022] It is evident that by using different filtering rules, target vehicles that pose a safety risk to autonomous driving can be accurately identified from environmental data. Further analysis based on the target vehicle's state information can reduce the amount of data required, improve processing efficiency, and ensure the driving safety of autonomous vehicles.

[0023] Optionally, determining the driving strategy of the autonomous vehicle based on the expected oncoming traffic situation further includes: when there is an obstacle with a speed less than a second speed threshold in the second driving trajectory and the distance between the obstacle and the autonomous vehicle is less than a second set distance threshold, determining the driving strategy of the autonomous vehicle to be braking to a stop; obtaining the driving coverage area corresponding to the second driving trajectory based on the shape parameters of the autonomous vehicle and the set body extension, and determining the driving strategy of the autonomous vehicle to be braking to a stop when the driving coverage area partially overlaps with the first driving trajectory of the target vehicle; if the distance between the target vehicle and the autonomous vehicle is less than a third set distance threshold, determining the driving strategy of the autonomous vehicle to be braking based on a set deceleration.

[0024] It is evident that by adopting deceleration, braking, or stopping strategies under different special circumstances, the driving safety of autonomous vehicles in various environments can be effectively guaranteed.

[0025] Secondly, this disclosure provides a vehicle intersection decision-making device, which includes:

[0026] The environmental recognition module is used to determine, in response to the acquired environmental data, whether the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0027] The target recognition module is used to determine the target vehicle to be analyzed and the target vehicle's state information from environmental data based on the set filtering rules. The state information includes the target vehicle's first driving trajectory and the shape parameters of the set parts.

[0028] The analysis module is used to determine the expected meeting state between the autonomous vehicle and the target vehicle based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the set parts of the target vehicle.

[0029] The processing module is used to determine the driving strategy of the autonomous vehicle based on the expected oncoming traffic situation.

[0030] Optionally, the analysis module is specifically used to: determine the target structure with the largest width of the target vehicle as the set part based on the type of the target vehicle; generate a trajectory expansion circle corresponding to the first driving trajectory with a point on the first driving trajectory as the center, based on the width of the target structure and the set expansion width, wherein the set expansion width is determined based on the type of the target vehicle; and determine the expected meeting state based on the intersection of the second driving trajectory and the trajectory expansion circle.

[0031] Optionally, the analysis module is specifically used to: determine the intersection point of the second driving trajectory and the first driving trajectory; determine the tangent point of the trajectory expansion circle and the second driving trajectory; determine the point closest to the autonomous vehicle among the intersection point and the tangent point as the conflict point; and determine the expected meeting state based on the target vehicle's action state, the autonomous vehicle's action state, and the conflict point.

[0032] Optionally, the analysis module is specifically used to: if the target vehicle's movement state includes a first speed and the autonomous vehicle's movement state includes a second speed, determine the first moment when the head of the autonomous vehicle reaches the conflict point based on the second speed; determine the second moment when the head of the target vehicle reaches the conflict point and the third moment when the rear of the vehicle reaches the conflict point based on the first speed; if the difference between the second moment and the first moment is greater than a first time threshold, determine the expected passing state as the autonomous vehicle passing the conflict point first without colliding with the target vehicle; or, if the difference between the first moment and the third moment is greater than a second time threshold, determine the expected passing state as the target vehicle passing the conflict point first without colliding with the autonomous vehicle; or, if the difference between the first moment and the second moment is less than a first time threshold, determine the expected passing state as a collision between the target vehicle and the autonomous vehicle.

[0033] Optionally, the processing module is specifically used to: if the expected oncoming traffic situation is that the autonomous vehicle and the target vehicle will not collide, determine the driving strategy to maintain the current driving state; if the expected oncoming traffic situation is that the target vehicle and the autonomous vehicle will collide, determine the driving strategy of the autonomous vehicle based on the first speed and the first driving trajectory of the target vehicle.

[0034] Optionally, the processing module is specifically used to: if the first speed is higher than the set first speed threshold, adjust the driving speed based on the set deceleration parameters until the target vehicle passes the conflict point first and there is no collision with the autonomous vehicle; if the first speed is lower than the set first speed threshold, determine the safe meeting time based on the distance between the target vehicle and the conflict point and the first speed; and redetermine the driving speed based on the location of the autonomous vehicle and the safe meeting time.

[0035] Optionally, the target recognition module is specifically used to: filter vehicles whose distance from the autonomous vehicle is greater than a first set distance threshold; filter vehicles located behind the autonomous vehicle identified in the environmental data based on the autonomous vehicle's driving direction; if the autonomous vehicle's driving direction is straight, filter vehicles in the same direction as the autonomous vehicle identified in the environmental data; or, if the autonomous vehicle's driving direction is left-turning, filter vehicles in the same direction going straight or turning right identified in the environmental data located to the right of the autonomous vehicle; or, if the autonomous vehicle's driving direction is right-turning, filter vehicles in the same direction going straight or turning left identified in the environmental data located to the left of the autonomous vehicle; determine vehicles whose angle with the autonomous vehicle's driving direction is within a set angle range from the filtered vehicles as target vehicles to be analyzed; and obtain the state information of the target vehicles based on the environmental data.

[0036] Optionally, the processing module further includes: when there is an obstacle with a speed less than a second speed threshold in the second driving trajectory and the distance between the obstacle and the autonomous vehicle is less than a second set distance threshold, determining that the driving strategy of the autonomous vehicle is to brake to a stop; obtaining the driving coverage area corresponding to the second driving trajectory based on the shape parameters of the autonomous vehicle and the set body extension amount, and determining that the driving strategy of the autonomous vehicle is to brake to a stop when the driving coverage area partially overlaps with the first driving trajectory of the target vehicle; and determining that the driving strategy of the autonomous vehicle is to brake based on a set deceleration if the distance between the target vehicle and the autonomous vehicle is less than a third set distance threshold.

[0037] Thirdly, this disclosure also provides a control device, which includes:

[0038] At least one processor;

[0039] and memory that is communicatively connected to at least one processor;

[0040] The memory stores instructions that can be executed by at least one processor, which, when executed by at least one processor, cause the control device to perform a vehicle intersection decision method as described in any embodiment of the first aspect of this disclosure.

[0041] Fourthly, this disclosure also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement a vehicle intersection decision-making method as described in any of the first aspects of this disclosure.

[0042] Fifthly, this disclosure also provides a computer program product comprising computer execution instructions, which, when executed by a processor, are used to implement the vehicle intersection decision-making method as described in any embodiment corresponding to the first aspect of this disclosure. Attached Figure Description

[0043] Figure 1 This is an application scenario diagram of the vehicle intersection decision-making method provided in the embodiments of this disclosure;

[0044] Figure 2 A flowchart of a vehicle intersection decision-making method provided in one embodiment of this disclosure;

[0045] Figure 3a A flowchart of a vehicle intersection decision-making method provided in yet another embodiment of this disclosure;

[0046] Figure 3b for Figure 3a The flowchart of the method for determining the expected meeting state provided in the embodiment shown is as follows;

[0047] Figure 3c for Figure 3a A schematic diagram showing the positional relationships of tangent points, intersection points, and conflict points provided in the illustrated embodiment;

[0048] Figure 3d for Figure 3a The flowchart of the method for determining the expected meeting state based on the conflict point and vehicle driving state provided in the embodiment shown is as follows:

[0049] Figure 3e for Figure 3a The flowchart of the method for determining the driving strategy provided in the embodiment shown is as follows;

[0050] Figure 4a A flowchart of a vehicle intersection decision-making method provided in yet another embodiment of this disclosure;

[0051] Figure 4b for Figure 4a The illustrated embodiment provides a schematic diagram of a scenario where there are obstacles on the second driving trajectory;

[0052] Figure 4c for Figure 4a The illustrated embodiment shows a scenario where the driving coverage area of ​​the autonomous vehicle overlaps with the first driving trajectory.

[0053] Figure 5 A schematic diagram of the structure of a vehicle intersection decision-making device provided in yet another embodiment of this disclosure;

[0054] Figure 6 This is a schematic diagram of the structure of a control device provided in yet another embodiment of this disclosure. Detailed Implementation

[0055] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this application as detailed in the appended claims.

[0056] The technical solutions of the embodiments of this application and how the technical solutions of the embodiments of this application solve the above-mentioned technical problems are described in detail below with specific examples. 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 now be described with reference to the accompanying drawings.

[0057] With the emergence of diverse transportation demands, autonomous driving technology is gradually being applied more widely. Autonomous vehicles based on this technology can progressively perform various tasks, such as transporting people and goods in different environments. Among these tasks, how autonomous vehicles navigate intersections is a significant challenge. Whether in urban roads or enclosed areas, intersections are essential points of convergence, turning, and evacuation for human-driven vehicles, autonomous vehicles, and pedestrians. These intersections present numerous factors affecting vehicle safety, making decision-making difficult and prone to safety hazards.

[0058] Especially at intersections without traffic lights, the lack of traffic lights restricts the movement of vehicles and pedestrians in directions other than the direction in which autonomous vehicles are traveling, resulting in a complex environment for autonomous vehicles. The driving strategies used at intersections with traffic lights cannot guarantee the safety of unmanned driving measurements at intersections without traffic lights, posing a huge challenge to the safe and efficient passage of autonomous vehicles.

[0059] In existing technologies, autonomous vehicles primarily make decisions at intersections based on traffic lights. When the traffic lights indicate that driving is permitted, the vehicle determines its driving strategy by considering the status of vehicles on both sides of the intersection. For intersections without traffic lights, the driving strategy is determined using a learning-based method, requiring training to learn driving strategies for different environments. However, this method lacks stability. In practical applications, if unfamiliar environments arise during training, the reliability of the autonomous vehicle's decisions cannot be guaranteed, thus compromising driving safety.

[0060] To address the aforementioned issues, this application provides a vehicle intersection decision-making method. By using environmental data and preset filtering rules, it identifies specific target vehicles that require attention. Then, based on the target vehicle's state and its own driving state, it determines a driving strategy to adapt to different complex environments and ensure the safety and reliability of autonomous driving vehicles.

[0061] Figure 1 This is an application scenario diagram of the vehicle intersection decision-making method provided in the embodiments of this application. For example... Figure 1 As shown, in the intersection decision-making process, the autonomous vehicle 100 confirms its entry into the intersection by recognizing environmental data, then identifies the target vehicle 110, and determines whether it will collide with the target vehicle 110 based on the first driving trajectory 111 of the target vehicle 110 and its own second driving trajectory 101, thereby determining its own driving strategy for passing through the intersection.

[0062] It should be noted that, Figure 1 The scenario shown is illustrated using only one autonomous vehicle and one target vehicle as an example, but this application embodiment is not limited to this; that is, the number of autonomous vehicles and target vehicles can be arbitrary.

[0063] The vehicle intersection decision-making method provided in this application is described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0064] Figure 2 A flowchart illustrating a vehicle intersection decision-making method provided in one embodiment of this application. Figure 2 As shown, including but not limited to the following steps:

[0065] Step S201: In response to the acquired environmental data, determine that the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0066] Specifically, autonomous vehicles collect environmental data about their surroundings in real time. This environmental data can include image data collected by image sensors (such as cameras), GPS positioning data collected by positioning sensors, and distance data from surrounding objects to the autonomous vehicle collected by infrared or ultrasonic sensors. Through this data, the vehicle's location and its relative position to surrounding objects (such as vehicles, lanes, and pedestrians) can be determined.

[0067] Depending on the type of data, the coverage (or distance) of environmental data varies. Typically, environmental data needs to cover an area of ​​more than 20 meters around the vehicle to ensure that when a dangerous object (such as a child sitting on the ground or a small animal) is detected in the autonomous vehicle's path, braking measures can be taken in time to ensure driving safety.

[0068] Autonomous vehicles need to use environmental data to determine that they are at an intersection in order to switch to the logic for entering the intersection (to determine whether an indicator light signal has been detected).

[0069] If the indicator light signal can be detected, then a driving strategy based on the indicator light signal in the relevant technology can be used for driving.

[0070] However, if the autonomous vehicle fails to detect the indicator light signal, it will need to make subsequent decisions based on other environmental data.

[0071] Intersection recognition can be approached in different ways depending on the type of environmental data. For example, with GPS positioning data, the location of the autonomous vehicle can be combined with map data from positioning or mapping software (e.g., if the vehicle is located at an intersection on the map, recognition is complete). With image data, recognition can be performed directly based on captured images of the surrounding environment (e.g., if the ground features match those of an intersection, recognition is complete). With distance data, recognition can be performed based on acquired road surface features (e.g., if the road surface features match those of an intersection, recognition is complete). Other existing intersection recognition methods can also be used, or a combination of methods can be employed.

[0072] Step S202: Based on the set filtering rules, determine the target vehicle and its status information from the environmental data.

[0073] The status information includes the target vehicle's first driving trajectory and the shape parameters of the designated location.

[0074] Specifically, since there are no traffic lights at intersections, when an autonomous vehicle passes through an intersection, there may be vehicles and pedestrians traveling in different directions around the autonomous vehicle (whereas if there were traffic lights at the intersection, there would usually only be other vehicles around the autonomous vehicle, traveling in the same or opposite direction, and vehicles traveling in a different direction would usually not intersect with the autonomous vehicle). Therefore, it is necessary to identify data that may affect the safety of the vehicle from the surrounding vehicles and pedestrians based on the collected environmental data.

[0075] The rules for determining whether surrounding vehicles and pedestrians will affect the safety of vehicle operation are called filtering rules. Among them, when there are pedestrians around the autonomous vehicle (such as when the distance between the pedestrian and the autonomous vehicle's driving trajectory is less than the set safe distance value), the autonomous vehicle will stop directly or decelerate at the set deceleration without further analysis (because the actions of pedestrians are difficult to predict, and to ensure safety, it is best to decelerate directly to a stop, or decelerate until the distance between the pedestrian and the driving trajectory reaches the set safe distance value).

[0076] If there are other vehicles around the autonomous vehicle, they can be filtered based on factors such as the distance between the autonomous vehicle and other vehicles (vehicles that are too far away can be eliminated directly), relative speed (vehicles that are too slow relative to the autonomous vehicle will not meet it after passing through the intersection, such vehicles can also be eliminated directly), and relative direction of travel (vehicles that are not in the same direction relative to the autonomous vehicle will be eliminated directly).

[0077] The remaining vehicles after filtering are those that may affect the driving safety of autonomous vehicles, i.e., target vehicles. These vehicles need to be further analyzed to ensure the safety of autonomous vehicles.

[0078] To facilitate further analysis, it is necessary to collect the target vehicle's status information. This status information can typically be determined by combining image data and distance data, including vehicle type (different types such as large trucks, tricycles, and cars, as their predicted deceleration distances and shape characteristics differ), vehicle shape parameters (depending on the target vehicle type, the corresponding set parts for shape parameters vary; typically, one of the widest parts of the target vehicle is selected as the set part, and the width of the set part is the corresponding shape parameter), driving speed, relative position to the autonomous vehicle, and the predicted driving trajectory (i.e., the first driving trajectory), etc.

[0079] The first driving trajectory can be determined by comprehensively considering the location, speed, and direction of the target vehicle collected continuously. This method of determination can directly refer to relevant technologies and is not limited here.

[0080] Step S203: Based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the set parts of the target vehicle, determine the expected meeting state between the autonomous vehicle and the target vehicle.

[0081] Specifically, because different types of target vehicles have different sizes and shapes, the analysis of their encounters with autonomous vehicles will produce different results. For example, if the target vehicle is a long truck, the analysis of whether it will collide with an autonomous vehicle needs to consider its speed, length, and width. However, if the target vehicle is a mini electric vehicle, its length can be disregarded and only its width needs to be considered, because the length of such vehicles is very short and has little impact on the analysis results. Therefore, it is necessary to select a set part for analyzing the encounter based on different vehicle types, and to determine whether a collision will occur based on the shape parameters of the set part.

[0082] Since both the target vehicle and the autonomous vehicle are usually traveling at high speeds at intersections, the analysis of the oncoming traffic situation mainly involves selecting a set part (such as the part with the widest width) to generate a shape model of the target vehicle (such as a cuboid model that includes the target vehicle). Then, the analysis is conducted to determine whether the position and trajectory of this shape model intersect with the trajectory of the autonomous vehicle's shape model (i.e., the second driving trajectory). If there is an intersection, the analysis is conducted to determine whether the shape models of the two vehicles will partially overlap when either vehicle moves to the intersection point (i.e., the expected oncoming traffic situation). If there is an overlap, it can be assumed that the two vehicles will collide. It is not necessary for the specific structures of the two vehicles to overlap, as this would be too computationally intensive and have little impact on the results.

[0083] By analyzing the shape parameters of a defined part, compared to related technologies that use pre-configured vehicle models or collect detailed parameters of the target vehicle, the computational load can be significantly reduced while ensuring the effectiveness and reliability of the analysis.

[0084] Step S204: Determine the driving strategy of the autonomous vehicle based on the expected meeting situation.

[0085] Specifically, if the expected scenario is that the two vehicles will not collide, the driving strategy can be to maintain the current driving state.

[0086] If the anticipated oncoming traffic situation suggests a potential collision, different driving strategies need to be determined based on the specific oncoming traffic situation. For example, if the two vehicles pass through the intersection one after the other, the driving strategy can be set to decelerate until the anticipated oncoming traffic situation changes to one where a collision will not occur; if the two vehicles pass through the intersection simultaneously, the driving strategy can be set to brake to a stop.

[0087] In some embodiments, other driving strategies may be adopted in specific emergency situations. For example, if the target vehicle is continuously decelerating, the driving strategy may be to accelerate through within a set speed range; if the target vehicle is continuously accelerating, the driving strategy may be to decelerate or to fine-tune the driving route so that the target vehicle can pass through the intersection first (but deceleration strategy is generally preferred), and so on.

[0088] The vehicle intersection decision-making method provided in this application determines, in response to acquired environmental data, whether an autonomous vehicle has entered an intersection and no traffic light signal has been detected. Then, based on predefined filtering rules, it identifies the target vehicle and its state information from the environmental data. Next, based on the target vehicle's first driving trajectory, the shape parameters of a predefined location, and the positional relationship between the autonomous vehicle's second driving trajectory and the target vehicle, it determines the expected meeting state between the autonomous vehicle and the target vehicle. Finally, based on the expected meeting state, it determines the autonomous vehicle's driving strategy. Therefore, even at intersections without traffic lights, it can effectively identify target vehicles affecting the autonomous vehicle's driving. Furthermore, based on the target vehicle's driving state and structural characteristics, it can determine whether the autonomous vehicle needs to slow down or evade. This ensures that the autonomous vehicle can reliably guarantee driving safety at intersections without traffic lights. By filtering target vehicles and determining driving strategies accordingly, it ensures vehicle driving safety while minimizing the impact of environmental factors, thus guaranteeing the vehicle's efficiency in passing through intersections.

[0089] Figure 3a A flowchart illustrating a vehicle intersection decision-making method provided in yet another embodiment of this application. For example... Figure 3a As shown, the vehicle intersection decision-making method includes:

[0090] Step S301: In response to the acquired environmental data, determine that the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0091] Step S302: Based on the set filtering rules, determine the target vehicle and its status information from the environmental data.

[0092] The status information includes the target vehicle's initial driving trajectory and shape parameters.

[0093] Specifically, steps S301 to S302 and Figure 2 The corresponding steps in the illustrated embodiments are the same and will not be repeated here.

[0094] Step S303: Based on the type of the target vehicle, the target structure with the largest width of the target vehicle is determined as the designated part.

[0095] Specifically, in order to determine the expected meeting state between the target vehicle and the autonomous vehicle, it is necessary to determine whether there will be an intersection point between the two based on the shape parameters and first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle.

[0096] For target vehicles, since different types of target vehicles have different sizes, directly using all size data would involve a large amount of calculation. Furthermore, different types of target vehicles have different adjustability during driving (for example, heavy trucks usually maintain their primary driving trajectory when they do not encounter dangerous situations, making them highly predictable, while small tricycles may accelerate or change their primary driving trajectory when they see potential dangerous situations, making them less predictable). Directly using size data to calculate whether a collision will occur may differ from the actual situation (especially for vehicles with more agile handling, which may accelerate and change their primary driving trajectory even if the predicted primary and secondary driving trajectories do not intersect, resulting in the altered primary and secondary driving trajectories intersecting and thus creating a collision risk).

[0097] Therefore, by determining its maximum width and combining it with the set expansion width, and then combining it with the first driving trajectory to determine whether there is an intersection point, we can fully consider the situation where the first driving trajectory changes, maximize the safety of the meeting process, and thus ensure the safety of autonomous vehicles when passing through intersections.

[0098] The specific correspondence between the target vehicle type and the target structure can be configured according to the actual situation. For example, a truck can choose the cargo box as the target structure, with the width of the cargo box as the maximum width; a car can choose the front of the vehicle as the target structure (because the front and rear widths of a car are usually the same), with the width of the front of the vehicle as the maximum width; a tricycle can choose the rear of the vehicle as the target structure (usually the rear of the vehicle has two wheels and is the widest part); and a motorcycle can choose the front of the vehicle as the target structure.

[0099] Step S304: Using a point on the first driving trajectory as the center, generate a trajectory expansion circle corresponding to the first driving trajectory based on the width of the target structure and the set expansion width.

[0100] The expansion width is determined based on the type of the target vehicle.

[0101] Specifically, since the first and second driving trajectories often do not intersect, but when the first and second driving trajectories are close to each other, the target vehicle based on the first driving trajectory and the autonomous vehicle based on the second driving trajectory may collide (i.e., a common scrape accident). In this case, it is impossible to make a judgment based on the intersection of the first and second driving trajectories.

[0102] Therefore, by generating several trajectory expansion circles corresponding to the first driving trajectory (each point on the first driving trajectory corresponds to the center of a trajectory expansion circle), when the second driving trajectory is tangent to any trajectory expansion circle, it can be assumed that there may be a collision between the target vehicle and the autonomous vehicle.

[0103] The diameter of the trajectory expansion circle is determined based on the width of the target structure (i.e., the width of the target vehicle) and the set expansion width.

[0104] The expansion width is typically set to include the width of the autonomous vehicle itself and the width of the target vehicle.

[0105] In some embodiments, the expansion width is set to the sum of the product of the target vehicle's width and the expansion coefficient, and the width of the autonomous vehicle itself. The expansion coefficient is determined based on the target vehicle type; generally, the larger the expansion coefficient, the more agile the target vehicle (e.g., a minimum of 1.2 for large trucks and a maximum of 3 for motorcycles). By setting the expansion coefficient, the distance margin in the expected passing situation is increased. This fully considers the change in the target vehicle's first driving trajectory when it may be close to the autonomous vehicle, ensuring that even if the target vehicle adjusts its first driving trajectory, it will not collide with the autonomous vehicle if its second driving trajectory is not tangent to any trajectory expansion circle. (Conversely, if the first and second driving trajectories do not intersect, but the target vehicle suddenly adjusts its trajectory during driving, causing the first and second driving trajectories to intersect, an accident will occur if the original driving strategy is followed. However, if the analysis is based on the trajectory expansion circle, even if the target vehicle adjusts its trajectory, a larger safety margin can be provided).

[0106] Step S305: Determine the expected meeting state based on the intersection of the second driving trajectory and the trajectory expansion circle.

[0107] Specifically, since the second driving trajectory will inevitably be tangent to or intersect with a certain trajectory expansion circle when it intersects with the first driving trajectory, the expected meeting state can be analyzed based on the intersection point of the second driving trajectory and any trajectory expansion circle.

[0108] Furthermore, such as Figure 3b The diagram shown is a flowchart of a method for determining the expected meeting point, and its specific steps include:

[0109] Step S3051: Determine the intersection point of the second driving trajectory and the first driving trajectory.

[0110] Specifically, in one scenario, the second driving trajectory intersects with the first driving trajectory, in which case the corresponding intersection point is directly determined.

[0111] Step S3052: Determine the tangent point between the trajectory expansion circle and the second driving trajectory.

[0112] Specifically, when the second driving trajectory intersects the first driving trajectory, it will inevitably be tangent to either trajectory expansion circle, thus there will always be a point of tangency. There may be one or multiple points of tangency (when there are multiple points of tangency, it usually includes the intersection point, such as...). Figure 3c (As shown).

[0113] Another scenario is that the second driving trajectory does not intersect with the first driving trajectory, but the second driving trajectory is tangent to a trajectory expansion circle. In this case, the point of tangency can also be determined.

[0114] Step S3053: Determine the point closest to the autonomous vehicle among the intersection point and the tangent point as the conflict point.

[0115] Specifically, if both intersection and tangency points exist, the point closest to the autonomous vehicle is identified as the conflict point; if only tangency points exist, the tangency point is identified as the conflict point.

[0116] For example, such as Figure 3c As shown, it is a schematic diagram of the positional relationship between tangent points, intersection points, and conflict points. The second driving trajectory 310 of the autonomous vehicle 300 intersects the first driving trajectory 330 of the target vehicle 320 at a point 340, and the second driving trajectory 310 and the corresponding trajectory expansion circle 350 of the first driving trajectory 330 have a tangent point 360 (there are two tangent points 360 in the figure, which are tangent to the left side of the front of the target vehicle 320 and tangent to the right side of the front of the vehicle). Based on the distance between the intersection point 340, the tangent point 360 and the autonomous vehicle, the tangent point 360 closest to the autonomous vehicle 300 is determined as the conflict point.

[0117] Step S3054: Determine the expected meeting state based on the target vehicle's action state, the autonomous vehicle's action state, and the conflict point.

[0118] Specifically, the expected meeting state between the target vehicle and the autonomous vehicle can be determined based on which vehicle arrives at the conflict point first, passes through the conflict point first, and whether the two vehicles will collide when they arrive at the conflict point.

[0119] In some embodiments, such as Figure 3d The diagram shows a flowchart of a method for determining the expected meeting state based on the conflict point and vehicle driving status. The specific steps include:

[0120] Step A1: Based on the second velocity, determine the first moment when the head of the autonomous vehicle reaches the point of conflict.

[0121] The target vehicle's operational status includes its first speed, while the autonomous vehicle's operational status includes its second speed.

[0122] Specifically, the size data and second speed of the autonomous vehicle can be directly determined, so the time when the autonomous vehicle (front part) arrives at the conflict point can be determined first, that is, the first moment.

[0123] Step A2: Based on the first speed, determine the second moment when the front of the target vehicle reaches the point of conflict and the third moment when the rear of the vehicle reaches the point of conflict.

[0124] Specifically, based on continuously acquired environmental data, the initial velocity of the target vehicle can be determined. Then, based on the length of the target vehicle, the second moment when its front reaches the point of conflict and the third moment when its rear reaches the point of conflict can be determined.

[0125] Step A3: If the difference between the second time point and the first time point is greater than the first time point threshold, determine the expected meeting state as the autonomous vehicle passing the conflict point first and not colliding with the target vehicle.

[0126] Specifically, if the second moment is shorter than the first moment and there is a certain difference (i.e., the first time threshold, such as 3 seconds), then when the autonomous vehicle passes the conflict point, the front of the target vehicle is still a certain distance away from the conflict point (combined with the first speed, usually more than 20 meters), and at this time the target vehicle and the autonomous vehicle will not collide.

[0127] Step A4: If the difference between the first time point and the third time point is greater than the second time threshold, the expected meeting state is determined to be that the target vehicle passes the conflict point first and there will be no collision with the autonomous vehicle.

[0128] Specifically, if the third time interval is shorter than the first time interval and there is a certain difference (i.e., the second time threshold, such as 2 seconds), then when the target vehicle passes the conflict point, the front of the autonomous vehicle is still a certain distance away from the conflict point (combined with the second speed, usually more than 10 meters), and at this time the target vehicle and the autonomous vehicle will not collide.

[0129] Step A5: If the difference between the first time point and the second time point is less than the first time point threshold, the expected passage state is determined to be that a collision will occur between the target vehicle and the autonomous vehicle.

[0130] Specifically, if the first moment is greater than the second moment, and the difference is small, the front of the target vehicle has already passed the collision point, but the rear has not, while the autonomous vehicle has also reached the collision point, resulting in the autonomous vehicle rear-ending the target vehicle. Conversely, if the first moment is less than the second moment, and the difference is small, the front of the autonomous vehicle has already passed the collision point, but the rear may not have, while the front of the target vehicle has also reached the collision point, resulting in the target vehicle rear-ending the autonomous vehicle. In both scenarios, a collision between the target vehicle and the autonomous vehicle will occur.

[0131] Steps A3 to A5 are parallel optional steps, and those skilled in the art can select the corresponding steps to perform according to the actual situation.

[0132] Step S306: If the expected oncoming traffic situation is that the autonomous vehicle and the target vehicle will not collide, determine the driving strategy to maintain the current driving state.

[0133] Specifically, if the autonomous vehicle will not collide with the target vehicle, the current driving state of the autonomous vehicle (including the second speed and the second driving trajectory) can be guaranteed to be safe. Therefore, it can maintain the current driving state and continue driving.

[0134] Step S307: If the expected oncoming situation is that the target vehicle and the autonomous vehicle will collide, determine the driving strategy of the autonomous vehicle based on the first speed and first driving trajectory of the target vehicle.

[0135] Specifically, in situations where a collision may occur, the driving strategy of autonomous vehicles needs to be adjusted, and different adjustment methods can be used depending on the initial speed of the target vehicle.

[0136] Furthermore, such as Figure 3e The diagram shown is a flowchart of the method for determining the driving strategy, and its specific steps include:

[0137] Step S3071: If the first speed is higher than the set first speed threshold, then adjust the driving speed based on the set deceleration parameters until the target vehicle passes the conflict point first and there is no collision with the autonomous vehicle.

[0138] Specifically, if the initial speed is not very low (the initial speed threshold is low, such as 20 km / h), the autonomous vehicle should directly decelerate or stop to ensure driving safety. The deceleration can be performed according to set deceleration parameters (such as the rate of deceleration and deceleration time) to ensure smooth and safe driving.

[0139] Step S3072: If the first speed is lower than the set first speed threshold, then determine the safe meeting time based on the distance between the target vehicle and the conflict point and the first speed.

[0140] Specifically, if the initial speed is very low (e.g., the target vehicle is passing through the intersection at a very low speed), the autonomous vehicle can choose to slow down, stop and wait, or readjust its speed and drive out of the intersection ahead of time without waiting for the target vehicle.

[0141] The specific method for adjusting the driving speed can be based on the difference between the second moment (which can be determined by the distance between the target vehicle and the conflict point and the first speed) and a set safe time threshold (such as 3 seconds), as the safe meeting moment.

[0142] Step S3073: Based on the location of the autonomous vehicle and the safe meeting time, redetermine the driving speed.

[0143] Specifically, by aiming to ensure that the autonomous vehicle passes through the conflict point at the moment of safe encounter, the driving speed of the autonomous vehicle is re-determined, which can avoid collisions between the autonomous vehicle and the target vehicle, thereby ensuring driving safety.

[0144] The vehicle intersection decision-making method provided in this application determines the target vehicle based on environmental data, generates a trajectory expansion circle based on the target vehicle's state information, then determines the expected meeting state based on the relationship between the autonomous vehicle's second driving trajectory and the trajectory expansion circle, and finally determines a driving strategy based on the expected meeting state. This ensures that the autonomous vehicle will not collide with the target vehicle when passing through the intersection, thereby guaranteeing safety when passing through the intersection and improving the safety and reliability of the autonomous vehicle.

[0145] Figure 4a A flowchart illustrating a vehicle intersection decision-making method provided in yet another embodiment of this application. Figure 4a As shown, the vehicle intersection decision-making method provided in this embodiment includes the following steps:

[0146] Step S401: In response to the acquired environmental data, determine that the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0147] Specifically, this step is related to Figure 2 The content of step S201 in the illustrated embodiment is the same, and will not be repeated here.

[0148] Step S402: Filter out vehicles whose distance from the autonomous driving vehicle is greater than a first set distance threshold.

[0149] Specifically, after determining that an autonomous vehicle has entered an intersection, it will first detect vehicles in the environment based on environmental data, and then filter out vehicles that will not pose a danger, such as vehicles that are too far away from the autonomous vehicle (exceeding a first set distance threshold, such as 100 meters).

[0150] Step S403: Based on the driving direction of the autonomous vehicle, filter out vehicles identified in the environmental data that are located behind the autonomous vehicle.

[0151] Specifically, when passing through an intersection, vehicles behind an autonomous vehicle usually follow the autonomous vehicle's trajectory, so there is usually no need to consider the possibility of a collision between them.

[0152] Step S404: If the autonomous vehicle is traveling in a straight direction, then filter out vehicles in the same direction as the autonomous vehicle identified in the environmental data.

[0153] Specifically, if an autonomous vehicle is going straight through an intersection, it does not need to consider vehicles going straight in the same direction or in opposite directions in different lanes, because the trajectories of these vehicles and the autonomous vehicle are parallel to each other, and there is usually no need to consider the possibility of a collision.

[0154] Step S405: If the autonomous vehicle is traveling in a left-turn direction, then filter out vehicles traveling in the same direction or turning right that are located to the right of the autonomous vehicle and identified in the environmental data.

[0155] Specifically, when an autonomous vehicle turns left, the straight-going and right-turning vehicles on its right side usually do not intersect with its driving trajectory, so this need not be considered.

[0156] Step S406: If the autonomous vehicle is traveling in the direction of a right turn, then filter out vehicles traveling in the same direction or turning left that are located to the left of the autonomous vehicle and identified in the environmental data.

[0157] Specifically, similar to the case of an autonomous vehicle turning left, the straight-going and left-turning vehicles on its left side usually do not intersect with its driving trajectory, so they do not need to be considered.

[0158] Steps S404 to S406 are parallel optional steps, and those skilled in the art can select the corresponding steps to perform according to the actual situation.

[0159] Step S407: Select vehicles from the filtered vehicles whose angle with the driving direction of the autonomous vehicle is within a set angle range as target vehicles to be analyzed.

[0160] Specifically, among the filtered vehicles, the relationship between their driving direction and that of the autonomous vehicle also needs to be considered. If the angle between the driving directions of the two vehicles is greater than a right angle (greater than 90 degrees or less than -90 degrees), it means that the two vehicles are gradually moving away from each other, so there is no need to consider the collision situation. However, if the angle between the driving directions on both sides is a right angle or less than a right angle (i.e., within the set angle range), there may be a collision risk, so it needs to be regarded as the target vehicle and further analyzed.

[0161] Step S408: Based on environmental data, obtain the status information of the target vehicle.

[0162] Specifically, after identifying the target vehicle, it is necessary to obtain its specific status information, such as shape parameters. For vehicles that are being filtered, it is not necessary to obtain this information in order to reduce the amount of data processing and improve processing efficiency.

[0163] Step S409: Based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the set parts of the target vehicle, determine the expected meeting state between the autonomous vehicle and the target vehicle.

[0164] Specifically, this step is related to Figure 2 and Figure 3a The corresponding steps in the illustrated embodiments are the same and will not be repeated here.

[0165] Step S410: Determine the driving strategy of the autonomous vehicle based on the expected meeting situation.

[0166] Specifically, after determining the expected meeting situation, the driving strategy of the autonomous vehicle can be further determined.

[0167] In some embodiments, if there is an obstacle in the second driving trajectory with a speed less than a second speed threshold and the distance between the obstacle and the autonomous vehicle is less than a second set distance threshold, the driving strategy of the autonomous vehicle is determined to be braking to a stop.

[0168] Specifically, if there are pedestrians, obstacles (such as roadblocks, gravel, etc.) or small animals (moving at a speed lower than the second speed threshold, such as 5m / s, in the second driving trajectory of the autonomous vehicle) within a relatively close distance (i.e., the second set distance threshold, such as 50 meters), it is necessary to brake to a stop until the pedestrians or small animals leave the second driving trajectory, or to replan the second driving trajectory.

[0169] like Figure 4b As shown, this is a schematic diagram of a scenario where there is an obstacle on the second driving trajectory. If there is an obstacle 410 on the driving trajectory of the autonomous vehicle 400, the autonomous vehicle needs to brake and stop in time.

[0170] In some embodiments, the driving coverage area corresponding to the second driving trajectory is obtained based on the shape parameters of the autonomous vehicle and the set body extension amount. When the driving coverage area partially overlaps with the first driving trajectory of the target vehicle, the driving strategy of the autonomous vehicle is determined to be braking to a stop.

[0171] Specifically, if the trajectory expansion circles of the autonomous vehicle and the target vehicle do not intersect or tangent, by extending the front of the autonomous vehicle forward by a predetermined body extension (e.g., 3 meters), the driving coverage area corresponding to the second driving trajectory (safe driving) can be obtained. Then, based on the driving coverage area, it is determined whether there is any overlap with the first driving trajectory or trajectory expansion circle of the target vehicle, so as to determine whether there is a sufficient safe distance when the vehicles are expected to meet, thereby maximizing the safety of the autonomous vehicle.

[0172] like Figure 4c As shown, it is a schematic diagram of a scenario where the driving coverage area of ​​the autonomous vehicle overlaps with the first driving trajectory. The driving coverage area 440 obtained by extending the driving trajectory of the autonomous vehicle 400 overlaps with the trajectory expansion circle 430 corresponding to the first driving trajectory of the target vehicle 420. At this time, the autonomous vehicle needs to brake and stop in time.

[0173] In some embodiments, if the distance between the target vehicle and the autonomous vehicle is less than a third preset distance threshold, the driving strategy of the autonomous vehicle is determined to be braking based on a preset deceleration.

[0174] Specifically, when the target vehicle and the autonomous vehicle are close to each other (i.e. less than the third set distance, such as less than 20 meters), the distance between the two vehicles is close regardless of the expected passing situation. At this time, the relative safety is low no matter how the vehicle is adjusted. Therefore, the autonomous vehicle needs to decelerate and brake directly to ensure safety.

[0175] The vehicle intersection decision-making method provided in this application requires, after acquiring environmental data, filtering out vehicles that do not affect the driving safety of autonomous vehicles according to different rules to obtain target vehicles that need further analysis. Then, based on the expected meeting state between the target vehicles and the autonomous vehicles, a driving strategy is determined. Thus, by filtering target vehicles and only acquiring their specific state information, the amount of information processing is reduced, processing efficiency is improved, and the driving safety of autonomous vehicles is ensured.

[0176] Figure 5 This is a schematic diagram of the structure of a vehicle intersection decision-making device provided in one embodiment of this application. Figure 5 As shown, the vehicle intersection decision-making device 500 includes: an environment recognition module 510, a target recognition module 520, an analysis module 530, and a processing module 540. Wherein:

[0177] The environment recognition module 510 is used to determine, in response to the acquired environmental data, that the autonomous vehicle has entered the intersection and no indicator light signal has been detected.

[0178] The target recognition module 520 is used to determine the target vehicle to be analyzed and the target vehicle's status information from environmental data based on the set filtering rules. The status information includes the target vehicle's first driving trajectory and the shape parameters of the set parts.

[0179] Analysis module 530 is used to determine the expected meeting state between the autonomous vehicle and the target vehicle based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of the set parts of the target vehicle.

[0180] The processing module 540 is used to determine the driving strategy of the autonomous vehicle based on the expected oncoming traffic situation.

[0181] Optionally, the analysis module 530 is specifically used to: determine the target structure with the largest width of the target vehicle as the set part based on the type of the target vehicle; generate a trajectory expansion circle corresponding to the first driving trajectory with a point on the first driving trajectory as the center, based on the width of the target structure and the set expansion width, wherein the set expansion width is determined based on the type of the target vehicle; and determine the expected meeting state based on the intersection of the second driving trajectory and the trajectory expansion circle.

[0182] Optionally, the analysis module 530 is specifically used to: determine the intersection point of the second driving trajectory and the first driving trajectory; determine the tangent point of the trajectory expansion circle and the second driving trajectory; determine the point closest to the autonomous vehicle among the intersection point and the tangent point as the conflict point; and determine the expected meeting state based on the target vehicle's action state, the autonomous vehicle's action state, and the conflict point.

[0183] Optionally, the analysis module 530 is specifically used to: if the target vehicle's movement state includes a first speed of the target vehicle and the autonomous vehicle's movement state includes a second speed of the autonomous vehicle, determine the first moment when the head of the autonomous vehicle reaches the conflict point based on the second speed; determine the second moment when the head of the target vehicle reaches the conflict point and the third moment when the rear of the vehicle reaches the conflict point based on the first speed; if the difference between the second moment and the first moment is greater than a first time threshold, determine the expected passing state as the autonomous vehicle passing the conflict point first and not colliding with the target vehicle; or, if the difference between the first moment and the third moment is greater than the second time threshold, determine the expected passing state as the target vehicle passing the conflict point first and not colliding with the autonomous vehicle; or, if the difference between the first moment and the second moment is less than the first time threshold, determine the expected passing state as a collision between the target vehicle and the autonomous vehicle.

[0184] Optionally, the processing module 540 is specifically used to: if the expected oncoming traffic situation is that the autonomous vehicle and the target vehicle will not collide, determine the driving strategy to maintain the current driving state; if the expected oncoming traffic situation is that the target vehicle and the autonomous vehicle will collide, determine the driving strategy of the autonomous vehicle based on the first speed and the first driving trajectory of the target vehicle.

[0185] Optionally, the processing module 540 is specifically used to: if the first speed is higher than the set first speed threshold, adjust the driving speed based on the set deceleration parameters until the target vehicle passes the conflict point first and there is no collision with the autonomous vehicle; if the first speed is lower than the set first speed threshold, determine the safe meeting time based on the distance between the target vehicle and the conflict point and the first speed; and redetermine the driving speed based on the location of the autonomous vehicle and the safe meeting time.

[0186] Optionally, the target recognition module 520 is specifically used to: filter vehicles whose distance from the autonomous vehicle is greater than a first set distance threshold; filter vehicles located behind the autonomous vehicle identified in the environmental data based on the autonomous vehicle's driving direction; if the autonomous vehicle's driving direction is straight, filter vehicles in the same direction as the autonomous vehicle identified in the environmental data; or, if the autonomous vehicle's driving direction is left-turning, filter vehicles in the same direction going straight or turning right identified on the right side of the autonomous vehicle identified in the environmental data; or, if the autonomous vehicle's driving direction is right-turning, filter vehicles in the same direction going straight or turning left identified on the left side of the autonomous vehicle identified in the environmental data; determine vehicles whose angle with the autonomous vehicle's driving direction is within a set angle range from the filtered vehicles as target vehicles to be analyzed; and obtain the state information of the target vehicles based on the environmental data.

[0187] Optionally, the processing module 540 further includes: when there is an obstacle with a speed less than a second speed threshold in the second driving trajectory and the distance between the obstacle and the autonomous vehicle is less than a second set distance threshold, determining that the driving strategy of the autonomous vehicle is to brake to a stop; obtaining the driving coverage area corresponding to the second driving trajectory based on the shape parameters of the autonomous vehicle and the set body extension amount, and determining that the driving strategy of the autonomous vehicle is to brake to a stop when the driving coverage area partially overlaps with the first driving trajectory of the target vehicle; and determining that the driving strategy of the autonomous vehicle is to brake based on a set deceleration if the distance between the target vehicle and the autonomous vehicle is less than a third set distance threshold.

[0188] In this embodiment, the vehicle intersection decision-making device, through the combination of various modules, enables autonomous vehicles to ensure driving safety with reliable rules even at intersections without traffic lights. Furthermore, by screening target vehicles and determining driving strategies accordingly, it ensures vehicle driving safety while being unaffected by excessive environmental factors, thus guaranteeing the efficiency of vehicle passage through intersections.

[0189] Figure 6 This is a schematic diagram of the structure of a control device provided in an embodiment of this application, as shown below. Figure 6 As shown, the control device 600 includes a memory 610 and a processor 620.

[0190] The memory 610 stores a computer program that can be executed by at least one processor 620. This computer program is executed by at least one processor 620 to enable the control device to implement the vehicle intersection decision-making method provided in any of the above embodiments.

[0191] The memory 610 and the processor 620 can be connected via a bus 630.

[0192] The relevant explanations can be understood by referring to the corresponding descriptions and effects in the method embodiments, and will not be repeated here.

[0193] One embodiment of this application provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the vehicle intersection decision method of any of the above embodiments.

[0194] The computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0195] One embodiment of this application provides a computer program product comprising computer-executable instructions that, when executed by a processor, are used to implement, as described above. Figures 2 to 4a The corresponding vehicle intersection decision method in any embodiment.

[0196] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0197] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope of this application is indicated by the claims.

[0198] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A vehicle intersection decision-making method, characterized in that, The vehicle intersection decision-making method includes: Based on the acquired environmental data, it is determined that the autonomous vehicle has entered the intersection and no indicator light signal has been detected. Based on the set filtering rules, the target vehicle to be analyzed and the state information of the target vehicle are determined from the environmental data. The state information includes the first driving trajectory of the target vehicle and the shape parameters of the set parts. Based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of a designated part of the target vehicle, the expected meeting state between the autonomous vehicle and the target vehicle is determined. Based on the expected meeting situation, the driving strategy of the autonomous vehicle is determined; The step of determining the expected meeting state between the autonomous vehicle and the target vehicle based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of a designated part of the target vehicle, includes: Based on the type of the target vehicle, the target structure with the largest width of the target vehicle is determined as the designated part; Using a point on the first driving trajectory as the center, and based on the width of the target structure and a set expansion width, generate a trajectory expansion circle corresponding to the first driving trajectory, wherein the set expansion width is determined based on the type of the target vehicle; Determine the intersection point of the second driving trajectory and the first driving trajectory; Determine the point of tangency between the trajectory expansion circle and the second driving trajectory; The point closest to the autonomous vehicle among the intersection point and the tangent point is identified as the conflict point; Based on the target vehicle's movement status, the autonomous vehicle's movement status, and the conflict point, the expected meeting state is determined.

2. The vehicle intersection decision-making method according to claim 1, characterized in that, The target vehicle's operational state includes a first speed, and the autonomous vehicle's operational state includes a second speed. Determining the expected meeting state based on the target vehicle's movement state, the autonomous vehicle's movement state, and the conflict point includes: Based on the second speed, determine the first moment when the head of the autonomous vehicle reaches the point of conflict; Based on the first speed, determine the second moment when the front of the target vehicle reaches the point of conflict and the third moment when the rear of the vehicle reaches the point of conflict; If the difference between the second time point and the first time point is greater than the first time point threshold, the expected meeting state is determined to be that the autonomous vehicle passes the conflict point first and will not collide with the target vehicle; or, If the difference between the first and third time points is greater than the second time threshold, the expected meeting state is determined to be that the target vehicle passes the conflict point first and there will be no collision with the autonomous vehicle; or, If the difference between the first and second moments is less than the first time threshold, the expected state is determined to be a collision between the target vehicle and the autonomous vehicle.

3. The vehicle intersection decision-making method according to claim 1, characterized in that, The step of determining the driving strategy for the autonomous vehicle based on the expected meeting point includes: If the expected oncoming situation is that the autonomous vehicle and the target vehicle will not collide, the driving strategy is to maintain the current driving state. If a collision is expected between the target vehicle and the autonomous vehicle, the driving strategy of the autonomous vehicle is determined based on the target vehicle's first speed and first driving trajectory.

4. The vehicle intersection decision-making method according to claim 3, characterized in that, If a collision is anticipated between the target vehicle and the autonomous vehicle, the driving strategy of the autonomous vehicle is determined based on the target vehicle's first speed and first trajectory, including: If the first speed is higher than the set first speed threshold, the driving speed is adjusted based on the set deceleration parameters until the target vehicle passes the conflict point first and there is no collision with the autonomous vehicle. If the first speed is lower than the set first speed threshold, then a safe meeting time is determined based on the distance between the target vehicle and the conflict point and the first speed; The driving speed is re-determined based on the location of the autonomous vehicle and the safe meeting point.

5. The vehicle intersection decision-making method according to any one of claims 1 to 4, characterized in that, The process of determining the target vehicle and its status information from environmental data based on the set filtering rules includes: Filter out vehicles whose distance from autonomous vehicles exceeds a first set distance threshold; Based on the driving direction of the autonomous vehicle, filter out vehicles located behind the autonomous vehicle identified in the environmental data. If the autonomous vehicle is traveling in a straight line, then filter out vehicles traveling in the same direction as the autonomous vehicle identified in the environmental data; or, If the autonomous vehicle is turning left, then filter out vehicles traveling straight or turning right to the right of the autonomous vehicle identified in the environmental data; or, If the autonomous vehicle is traveling in the direction of a right turn, then filter out vehicles traveling in the same direction straight or turning left that are located to the left of the autonomous vehicle and identified in the environmental data. Vehicles that are filtered and whose angle with the autonomous vehicle's direction of travel is within a set range are identified as target vehicles to be analyzed. Based on the environmental data, the status information of the target vehicle is obtained.

6. The vehicle intersection decision-making method according to any one of claims 1 to 4, characterized in that, Based on the expected meeting point, determining the driving strategy for the autonomous vehicle further includes: When there is an obstacle with a speed less than a second speed threshold in the second driving trajectory and the distance between the obstacle and the autonomous vehicle is less than a second set distance threshold, the driving strategy of the autonomous vehicle is determined to be braking to a stop. Based on the shape parameters of the autonomous vehicle and the set body extension, the driving coverage area corresponding to the second driving trajectory is obtained. When the driving coverage area partially overlaps with the first driving trajectory of the target vehicle, the driving strategy of the autonomous vehicle is determined to be braking to a stop. If the distance between the target vehicle and the autonomous vehicle is less than a third set distance threshold, the autonomous vehicle's driving strategy is determined to be braking based on a set deceleration.

7. A vehicle intersection decision-making device, characterized in that, include: The environmental recognition module is used to determine, in response to the acquired environmental data, whether the autonomous vehicle has entered the intersection and no indicator light signal has been detected. The target recognition module is used to determine the target vehicle to be analyzed and the target vehicle's state information from environmental data based on the set filtering rules. The state information includes the target vehicle's first driving trajectory and the shape parameters of the set parts. The analysis module is used to determine the expected meeting state between the autonomous vehicle and the target vehicle based on the positional relationship between the first driving trajectory of the target vehicle and the second driving trajectory of the autonomous vehicle, and the shape parameters of a set part of the target vehicle. The processing module is used to determine the driving strategy of the autonomous vehicle based on the expected meeting state; The analysis module is specifically used for: Based on the type of the target vehicle, the target structure with the largest width of the target vehicle is determined as the designated part; Using a point on the first driving trajectory as the center, and based on the width of the target structure and a set expansion width, generate a trajectory expansion circle corresponding to the first driving trajectory, wherein the set expansion width is determined based on the type of the target vehicle; Determine the intersection point of the second driving trajectory and the first driving trajectory; Determine the point of tangency between the trajectory expansion circle and the second driving trajectory; The point closest to the autonomous vehicle among the intersection point and the tangent point is identified as the conflict point; Based on the target vehicle's movement status, the autonomous vehicle's movement status, and the conflict point, the expected meeting state is determined.

8. A control device, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, cause the control device to perform the vehicle intersection decision method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the vehicle intersection decision-making method as described in any one of claims 1-6.

10. A computer program product, characterized in that, The program product includes computer-executable instructions, which, when executed by a processor, are used to implement the vehicle intersection decision-making method as described in any one of claims 1-6.

Images