WO2022053026A1

WO2022053026A1 - Automatic driving meeting scene processing method and apparatus, vehicle, and storage medium

Info

Publication number: WO2022053026A1
Application number: PCT/CN2021/117764
Authority: WO
Inventors: 张双琳; 颜波; 徐成; 张放; 李晓飞; 张德兆; 王肖; 霍舒豪
Original assignee: 武汉智行者科技有限公司
Priority date: 2020-09-11
Filing date: 2021-09-10
Publication date: 2022-03-17
Also published as: CN112265542B; CN112265542A

Abstract

An automatic driving meeting scene processing method, comprising the steps: step S1, meeting state determination: determining whether a vehicle is in a meeting state or not, and continuing to execute step S2 when the vehicle is in a meeting state; step S2, meeting type determination and processing: classifying the current meeting state, determining a corresponding meeting type, and then executing a corresponding meeting processing operation according to preset corresponding relationships between a plurality of different meeting types and a plurality of different meeting processing operations. Also disclosed are an automatic driving meeting scene processing apparatus and a vehicle. According to the automatic driving meeting scene processing method and apparatus, the vehicle, and a storage medium, design is scientific, and the speed and the transverse position of the vehicle can be actively adjusted in a meeting scene, thereby reducing the speed of the vehicle at the moment of meeting, increasing the transverse spacing distance between the vehicle and an opposite vehicle, improving the driving safety, and satisfying driving habits of people.

Description

A kind of automatic driving meeting scene processing method and device, vehicle, storage medium

This application claims the priority of the Chinese patent application filed on September 11, 2020 with the application number 202010951662.2 and titled "A method for processing an autonomous driving meeting scene and its device and vehicle". The disclosure of the Chinese patent application The contents are incorporated herein by reference.

technical field

The present invention relates to the technical field of automatic driving, and in particular, to a method for processing a car meeting scene of automatic driving, a device, a vehicle and a storage medium thereof.

Background technique

As the field of application of autonomous vehicles expands, so does the number of driving scenarios that need to be addressed. Meeting cars is a common traffic scene in the process of driving. In the traffic rules, many systems and regulations are proposed for meeting cars.

Meeting vehicles means that vehicles traveling in the opposite direction (including vehicles and opposite vehicles) pass through staggeredly at the same place at the same time. During the meeting, the vehicle speed should be properly controlled according to the conditions of the road, climate and vehicle type, sufficient horizontal distance should be maintained, and the relevant regulations of the meeting should be strictly observed. Due to the certain collision risk in some meeting conditions, it will cause human drivers to panic in psychology and operation.

At present, the existing solutions for obstacles in the field of autonomous driving can be roughly divided into two types: sampling and optimization. Among them, the sampling-based obstacle processing scheme searches for a guiding blank area by performing collision detection on the sampling points in the state space, so as to find a collision-free planning path from the starting point to the target point. Based on the optimized obstacle handling scheme, the behavior of obstacles is firstly determined, and the unknown parameters of kinematics are solved with safety, comfort and efficiency as the optimization goals.

However, the sampling-based obstacle processing scheme is not accurate enough to describe the environmental information, and lacks flexibility in selecting the search end point. However, the optimization-based obstacle processing scheme requires a large amount of computation and may not converge.

In addition, neither of the two existing obstacle handling solutions has a strategy for adjusting the characteristics of the car meeting scene, which is not in line with human driving habits.

Therefore, there is an urgent need to develop a method for dealing with car meeting scenarios that is safe and conforms to people's driving habits, which will be of great significance.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method for processing a car meeting scene of an automatic driving, a device, a vehicle and a storage medium thereof, aiming at the technical defects existing in the prior art.

To this end, the present invention provides a method for processing an automatic driving meeting scene, comprising the following steps:

Step S1: Judgment of the meeting state: judging whether the vehicle is in the meeting state, and when the vehicle is in the meeting state, proceed to step S2;

Step S2, judging processing of the meeting vehicle type: classifying the current meeting vehicle status, judging the corresponding meeting vehicle type, and then according to the preset correspondence between multiple different meeting vehicle types and multiple different meeting processing operations, Perform the corresponding meeting handling operation.

Preferably, after step S2, the following steps are also included:

Step S3, according to the type of meeting vehicle to which the current meeting state determined in step S2 belongs, determine the lateral rightward offset distance and target speed of the vehicle;

Step S4, according to the lateral rightward offset distance and the target speed of the vehicle determined in step S3, the planned target end point of the vehicle is obtained, and the planned target end point of the vehicle is the end point of the reference path of the vehicle;

Step S5, according to the current position of the vehicle and the position of the planned target end point of the vehicle determined in step S4, generate a driving track between the vehicle from the current position to the planned target end point.

Preferably, in step S11, it is first judged whether the vehicle meets the preset meeting conditions, and if it meets the preset meeting conditions, step S12 is performed; otherwise, it is judged that the vehicle is not in the meeting state, and no processing is performed;

The preset vehicle meeting conditions are specifically: the vehicle is located in the leftmost lane, and there is no physical obstruction on the left side of the vehicle;

Step S12, screening all the opposite vehicles in front of the vehicle (that is, vehicles driving in the opposite direction), if there is a situation where the opposite vehicle is driving in the adjacent opposite lane of the vehicle, it is judged that the vehicle is in a meeting state at this time, and step S2 is continued, Otherwise, it is judged that the vehicle is not in the meeting state, and no processing is performed.

Preferably, step S2 specifically includes the following sub-steps:

Step S21, first, it is judged whether there is a collision risk between the predicted trajectory of the opposite vehicle and the reference path of the vehicle, if there is no collision risk, it is judged that the current meeting state belongs to the speed limit type, and the corresponding meeting processing operation is: reduce the vehicle's collision risk. speed, and if there is a risk of collision, it is determined that the current meeting state does not belong to the speed limit type, and step S22 is continued;

Step S22, determine whether there is a conflict area between the vehicle and the opposite vehicle (that is, an area where two vehicles are not allowed to pass at the same time), if there is no conflict area, it is judged that the current meeting vehicle state belongs to the avoidance type, and the corresponding vehicle meeting processing operation is executed. It is: avoid collision by controlling the vehicle to move laterally to the right, and if there is a conflict area, it is judged that the current meeting state belongs to the parking type, and the corresponding meeting processing operation is: control the vehicle to stop before the conflict area, Wait for the oncoming vehicle to pass.

Preferably, in step S21, in order to determine whether there is a collision risk between the predicted trajectory of the opposite vehicle and the reference path of the vehicle, the specific operations are as follows:

Calculate the distance between any position point of the predicted trajectory of the opposite vehicle and any position point in the reference path of the vehicle, if there is a situation where the distance value is less than the preset safety distance, it means that there is a collision risk, otherwise it does not exist;

In step S22, the determination of whether there is a conflict area between the vehicle and the opposite vehicle specifically includes the following operations:

When there is a static obstacle in the opposite lane, and the width of the remaining space on the right side of the static obstacle on the opposite lane is less than the sum of the vehicle width, the width of the opposite vehicle and the preset safety distance, then it is judged that the vehicle is in the opposite direction. There is a conflict area between the vehicles.

Preferably, in step S3, according to the type of meeting vehicle to which the current meeting state determined in step S2 belongs, the lateral rightward offset distance of the vehicle is determined, which specifically includes the following steps:

If the current meeting state is of the speed limit type, the vehicle's lateral right offset distance is zero;

If the current meeting vehicle state is an avoidance type, calculate the lateral rightward offset distance of the vehicle according to the following calculation formula. The specific formula is as follows:

s=l+0.5w+d;

Wherein, l represents the straight-line distance between the rightmost point of the predicted trajectory of the opposite vehicle and the reference path of the vehicle;

Among them, w represents the width of the vehicle; d represents the preset safety distance, that is, the horizontal vertical distance maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle;

If the current meeting state belongs to the parking type, calculate the lateral rightward offset distance of the vehicle according to the following calculation formula. The specific formula is as follows:

s=x-d;

Among them, x represents the distance from the center of the vehicle to the far right side of the road in front of the conflict area;

d represents the preset safety distance, that is, a certain horizontal and vertical distance is maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle.

Preferably, in step S3, the target speed of the vehicle is determined according to the type of meeting vehicle to which the current meeting state belongs, which specifically includes the following steps:

If the current meeting state is of the speed limit type, the target speed of the vehicle is the preset first target speed;

If the current meeting state is an avoidance type, the target speed of the vehicle is the preset second target speed;

Wherein, the preset second target speed value is smaller than the preset first target speed value;

If the current meeting state is of the parking type, the target speed of the vehicle is zero.

Preferably, in step S3, the following steps are also included:

Control the vehicle to perform a uniform deceleration motion, and calculate the deceleration distance for the vehicle to reach the meeting position. The specific formula is as follows:

Among them, v _ego represents the current speed of the vehicle;

v _goal represents the target speed of the vehicle;

v _obj represents the current speed of the opposite vehicle;

y represents the distance between the current position of the vehicle and the opposite vehicle.

In addition, the present invention also provides an automatic driving meeting scene processing device, including the following modules:

The meeting vehicle state judgment module is used to judge whether the vehicle is in the meeting vehicle state. When the vehicle is in the meeting vehicle state, it sends a trigger signal to the meeting vehicle judgment module to trigger the operation of the meeting vehicle type judgment module;

The meeting vehicle type judging processing module is connected with the meeting vehicle state judging module, and is used to classify the current meeting vehicle state after being triggered to run by the meeting vehicle state judging module, judge the corresponding meeting vehicle type, and then according to the preset The corresponding relationship between the multiple different meeting vehicle types and the multiple different meeting processing operations is performed, and the corresponding meeting processing operation is performed.

Preferably, the following modules are also included:

The lateral right offset distance and the target speed determination module are connected with the meeting vehicle type judgment processing module, and are used for determining the lateral direction of the vehicle according to the type of meeting vehicle to which the current meeting vehicle state determined by the meeting vehicle type judgment processing module belongs. Right offset distance and target speed;

The planning target end point determination module is connected with the lateral right offset distance and the target speed determination module, and is used for determining the lateral right offset distance and the target speed of the vehicle according to the lateral right offset distance and the target speed determination module. , obtain the planned target end point of the vehicle, and the planned target end point of the vehicle is the end point of the reference path of the vehicle;

The driving planning module is connected with the planning target end point determination module, and is used for generating the distance between the vehicle from the current position to the planning target end point according to the current position of the vehicle and the position of the planning target end point of the vehicle determined by the planning target end point determination module. driving track.

In addition, the present invention also provides a vehicle, including the above-mentioned apparatus for processing a car meeting scene of an automatic driving.

In addition, the present invention also provides a computer-readable storage medium, where a program is stored in the computer-readable storage medium, and when the program is executed by a multi-core processor, the multi-core processor causes the multi-core processor to execute the automatic The processing method of the driving meeting scene.

It can be seen from the above technical solutions provided by the present invention that, compared with the prior art, the present invention provides an automatic driving meeting scene processing method and device, vehicle, and storage medium, which are scientifically designed and can be used when encountering a meeting scene. At the same time, it can actively adjust the speed and lateral position of the vehicle, thereby reducing the speed of the vehicle at the moment of passing, and at the same time increasing the lateral separation distance between the vehicle and the opposite vehicle, which is conducive to further improving the safety of driving and satisfying people's driving habits , has great practical significance.

Description of drawings

Fig. 1 is a basic flowchart of a method for processing a car meeting scene of an automatic driving provided by the present invention;

2 is a schematic flowchart of an embodiment of a method for processing a car meeting scenario provided by the present invention;

FIG. 3 is a schematic diagram of a speed-limited type meeting scene faced by the present invention;

4 is a schematic diagram of an avoidance type meeting scene faced by the present invention;

FIG. 5 is a schematic diagram of a parking type meeting scene faced by the present invention.

detailed description

In order to make the technical means realized by the present invention easier to understand, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the related application, but not to limit the application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings.

It should be noted that the embodiments in the present application and the features of the embodiments may be combined with each other in the case of no conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

It should be noted that in the description of this application, the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. The terminology of the indicated direction or positional relationship is based on the direction or positional relationship shown in the drawings, which is only for convenience of description and does not indicate or imply that the device or element must have a particular orientation, be constructed and operated in a particular orientation , so it cannot be construed as a limitation on this application.

Referring to FIGS. 1 to 5 , the present invention provides a method for processing a car meeting scenario of automatic driving, including the following steps:

In the present invention, in terms of specific implementation, the operation of judging the meeting state of the vehicle in step S1 specifically includes the following sub-steps:

Step S11, first determine whether the vehicle meets the preset meeting conditions, if it meets the preset meeting conditions, then execute step S12, otherwise, it is judged that the vehicle is not in the meeting state, and no processing is performed;

The preset meeting conditions are specifically: the vehicle is located in the leftmost lane, and there is no physical obstruction (ie, an obstacle, such as a physical obstruction such as a guardrail) on the left side of the vehicle; both conditions need to be satisfied at the same time.

In step S11, it should be noted that the self-driving vehicle relies on the existing high-precision map in the self-driving state, and the high-precision map pre-stores information such as lane type, lane topology, lane line and geographic location.

According to the output of the positioning module installed on the vehicle, the absolute position of the vehicle can be obtained, and then the lane where the vehicle is located can be obtained based on the high-precision map. The vehicle is in the leftmost lane.

Among them, if the vehicle belongs to the leftmost lane, it is a prerequisite for the meeting. If there is a parallel lane on the left side of the vehicle, and the vehicle on the left is in the same direction as the vehicle, it does not constitute a meeting scene.

That is to say, judging whether the vehicle is in the leftmost lane is an existing mature technology, which will not be repeated here.

It should also be noted that the existing high-precision map pre-records the left and right boundary types, solid lines, dashed lines or physical occlusions of the lane. If the left boundary type of the vehicle lane is not a physical occlusion, it means that the left side of the vehicle does not exist. Physical Obstruction.

Among them, the absence of physical obstructions on the left side of the vehicle is a prerequisite for the meeting. If there is a physical obstruction by the guardrail light on the left side, there will be no vehicles on the left side that move in the opposite direction to the vehicle, which does not constitute a meeting scene.

Similarly, judging whether there is a physical obstruction on the left side of the vehicle is an existing mature technology and will not be repeated here.

Step S12, screening all the opposite vehicles in front of the vehicle (that is, vehicles driving in the opposite direction), if there is a situation where the opposite vehicle is driving in the adjacent opposite lane of the vehicle, it is judged that the vehicle is in a meeting state at this time, and step S2 is continued, Otherwise, it is judged that the vehicle is not in a meeting state, and no processing is performed;

In step S12, it should be noted that, according to the output of the existing perception module (such as lidar, camera, etc.) set on the vehicle, the relative position of the obstacles (including the opposite vehicle) around the vehicle can be obtained, so as to screen the vehicle Information on all oncoming vehicles ahead (i.e. vehicles driving in the opposite direction), speed and category. Among them, the vehicle whose position is in front of the vehicle and whose speed direction is opposite to that of the vehicle belongs to the opposite vehicle.

In the present invention, the opposite vehicle in front of the vehicle is a necessary condition for the target vehicle in the meeting scene.

In the present invention, regarding whether there is a situation in which the opposite vehicle is driving in the adjacent opposite lane of the vehicle, specifically: through the existing perception module, the absolute position of the obstacle (including the opposite vehicle) around the vehicle can be obtained, and then Through the high-precision map, the corresponding lane can be obtained according to the absolute position of the obstacles around the vehicle, and all the lanes with topological relationship of each lane are provided at the same time (the high-precision map pre-stores information such as lane type, lane topology, lane line and geographic location, etc. ). If the adjacent opposite lane of the lane where the oncoming vehicle (as an obstacle around the vehicle) is located includes the vehicle lane, it means that the oncoming vehicle is driving in the opposite lane adjacent to the vehicle. This is an existing mature technical means, which will not be repeated here.

Among them, the opposite vehicle driving in the opposite lane adjacent to the vehicle is a necessary condition for the target vehicle in the meeting scene.

Step S2, judging processing of the meeting vehicle type: classifying the current meeting vehicle status, judging the corresponding meeting vehicle type, and then according to the preset correspondence between multiple different meeting vehicle types and multiple different meeting processing operations, Execute the corresponding car meeting processing operation;

In the present invention, in terms of specific implementation, step S2 specifically includes the following sub-steps:

It should be noted that when the current meeting state is of the speed limit type, the driver's sense of security can be improved by reducing the vehicle speed, as shown in Figure 3.

In the present invention, in terms of specific implementation, in step S21, the predicted trajectory of the opposite vehicle can be calculated according to the distribution of surrounding obstacles combined with the motion state, and is usually obtained by kinematic trajectory estimation or machine learning algorithm.

It should be noted that, regarding the predicted trajectory of the oncoming vehicle, the prediction module in the existing automatic driving system installed on the vehicle can estimate the oncoming vehicle based on the historical position and speed of obstacles around the vehicle such as the oncoming vehicle. The position information of obstacles around vehicles such as vehicles in a period of time in the future is to obtain the predicted trajectories of obstacles around vehicles such as opposing vehicles. At present, the predicted trajectories of obstacles around vehicles such as oncoming vehicles can be obtained through existing algorithms such as Kalman filtering or deep learning. This is an existing mature technology and will not be repeated here.

Wherein, the predicted trajectory of the oncoming vehicle may be, for example, the position information of the oncoming vehicle corresponding to each 0.1 second interval for the next 8 seconds. The predicted trajectory is an important basis for judging the type of meeting scene.

It should be noted that, with regard to the reference path of the vehicle, through the existing automatic driving planning module set on the vehicle, the position points that the vehicle needs to pass through in the future can be output in real time according to the environmental information around the vehicle. The combination of the position points is Reference path. At present, the reference path of the vehicle can be obtained through existing algorithms such as hybrid A* and quintic polynomial optimization. This is an existing mature technology and will not be repeated here.

The reference path of the vehicle can be the position and speed information of the vehicle corresponding to every 0.1 second interval in the next 8 seconds.

It should also be noted that the reference path of the vehicle is an important basis for judging the type and status of the meeting scene, and the position of the vehicle can be controlled through the reference path planning, thereby realizing the meeting function.

In step S21, in order to determine whether there is a collision risk between the predicted trajectory of the opposite vehicle and the reference path of the vehicle, the specific operation is as follows:

Calculate the distance between the predicted trajectory of the opposite vehicle and the reference path of the vehicle (between any two position points, that is, any position point of the predicted trajectory of the opposite vehicle and any position point in the reference path of the vehicle) If there is a situation where the distance value is less than the preset safety distance (for example, 5m), it means that there is a collision risk, otherwise it does not exist.

In the present invention, whether there is a collision between the predicted trajectory and the reference path is an important basis for judging the type of the meeting scene.

In the present invention, in step S22, judging whether there is a conflict area between the vehicle and the opposite vehicle specifically includes the following operations:

Referring to Figure 5, when the static obstacle on the left (that is, there is a static obstacle on the opposite lane) occupies more of the passable area, and the width of the remaining space on the right side of the static obstacle on the opposite lane is smaller than that of the vehicle The sum of the width, the width of the opposite vehicle and the preset safety distance (for example, 0.5m, that is, the reserved safety distance between the right side of the opposite vehicle and the left side of the vehicle), then it is judged that there is a conflict area between the vehicle and the opposite vehicle, The remaining space to the right of the static obstacle on the opposite lane (horizontal right direction space) belongs to the conflict area.

In terms of specific implementation, judging whether there is a conflict area between the vehicle and the opposite vehicle may specifically include the following sub-steps:

The first step is to perform detection at a certain distance (1m) between the vehicle and the opposite vehicle, and obtain the distance corresponding to the left road boundary point and the right boundary point of the detection distance from the existing high-precision map.

The second step is to screen all static obstacles that cover the detection distance, that is, obstacles whose closest distance to the vehicle is greater than the detection distance and the farthest distance to the vehicle is less than the detection distance. The obstacles are located at the left and right of the detection distance. Side boundary point and right boundary point.

The third step is to sort all obstacles in the order from left to right, and calculate the passable distance between adjacent obstacles, that is, the right boundary point of the left obstacle and the left boundary point of the right obstacle. Distance, as an alternative travelable distance. The distance between the left road boundary point and the left boundary point of the leftmost obstacle and the distance between the right boundary point of the rightmost obstacle and the right road boundary point are also used as alternative traversable distances.

The fourth step is to calculate the width of the opposite vehicle, the sum of the width of the vehicle and the left and right safety distances as the minimum simultaneous passing distance. If the minimum simultaneous passing distance is greater than the largest alternative passable distance, it indicates that there is a conflict. If any detected distances are in conflict, there is a conflict area between the vehicle and the oncoming vehicle.

In the present invention, judging whether there is a conflict area between the vehicle and the opposite vehicle is an important basis for judging the type of the meeting scene.

It should be noted that, in the present invention, the collision position is calculated as the end point of the vehicle decelerating and stopping, which can ensure that the vehicle does not obstruct the passage of the oncoming vehicle.

It should be noted that if there is no conflict area between the vehicle and the opposite vehicle, that is, there is no area where two vehicles are not allowed to pass at the same time, it means that the vehicle can avoid the collision by moving laterally, and the current state of meeting vehicles belongs to avoidance. type, as shown in Figure 4. Otherwise, the current meeting state belongs to the parking type, and it is necessary to stop before the conflict area and wait for the oncoming vehicle to pass, as shown in Figure 5.

In the present invention, after step S2, the following steps may also be included:

Step S4, obtaining the planned target end point of the vehicle (that is, the end point of the reference path of the vehicle) according to the lateral rightward offset distance and the target speed of the vehicle determined in step S3;

In the present invention, it should be noted that, regarding step S4, the reference path of the vehicle and the end point of the reference path can be obtained through existing algorithms such as hybrid A* and quintic polynomial optimization. This is an existing mature technology and will not be repeated here.

Step S5, according to the current position of the vehicle and the position of the planned target end point of the vehicle (that is, the end point of the reference path of the vehicle) determined in step S4, generate a driving trajectory between the vehicle from the current position to the planned target end point (for example, using existing search algorithms or other existing algorithms capable of generating driving paths).

It should be noted that the driving trajectory is a trajectory that satisfies the vehicle dynamics constraints and is safe.

In the present invention, it should be noted that, with regard to step S5, existing algorithms such as hybrid A* and quintic polynomial optimization can be used, according to the current position of the vehicle and the planned target end point of the vehicle determined in step S4 (that is, the vehicle The position of the end point of the reference path), and the driving trajectory between the vehicle from the current position to the end point of the planned target is generated. This is an existing mature technology and will not be repeated here.

It should be noted that, for the present invention, with regard to step S3, in order to ensure that there is a sufficient safety distance during the meeting of vehicles, the corresponding lateral offset distance needs to be calculated according to the type of the meeting vehicles. If the current speed limit type, the vehicle can maintain the current reference path, and does not need to offset to the right. If it is currently in the avoidance type, the vehicle needs to maintain a safe distance from the far right side of the obstacle.

In the present invention, in terms of specific implementation, in step S3, according to the type of meeting vehicle to which the current meeting state determined in step S2 belongs, the lateral rightward offset distance of the vehicle is determined, which specifically includes the following steps:

If the current meeting state is of the speed limit type, the vehicle does not need to be offset laterally to the right, that is, the lateral offset distance of the vehicle to the right is zero.

s=l+0.5w+d;

Among them, l represents the straight-line distance between the rightmost point of the predicted trajectory of the opposite vehicle (which belongs to the dynamic obstacle) and the reference path of the vehicle (considering the obstacle area), and the right is positive;

Among them, w represents the width of the vehicle, and d represents the preset safety distance (that is, the horizontal vertical distance maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle). Considering the perception and control errors, d is generally taken as 0.5 About m, for example, 0.5m.

It should be noted that the lateral right offset distance of the vehicle refers to the distance that the waypoint of the vehicle reference path is shifted to the right along the road direction.

In terms of specific implementation, l represents the straight-line distance between the rightmost point of the predicted trajectory of the opposite vehicle (which belongs to a dynamic obstacle) and the reference path of the vehicle (considering the area of the obstacle), specifically: by calculating the prediction of the opposite vehicle The four corner points on the opposite vehicle (that is, the four corner position points of the vehicle) corresponding to each point in the trajectory, and the vertical distance from the vehicle reference path, take the minimum value, and the minimum value is taken as l.

Here, it should be noted that if the current passing vehicle state belongs to the avoidance type, the vehicle needs to maintain a certain safety distance d from the opposite vehicle (which belongs to the dynamic obstacle) on the far right side.

s=x-d;

Among them, x represents the distance from the center of the vehicle to the rightmost boundary line of the road in front of the conflict area;

d represents the preset safety distance (that is, a certain horizontal and vertical distance is maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle). Considering the perception and control errors, d is generally about 0.5m, for example, 0.5m .

In terms of specific implementation, according to the existing high-precision map, the right boundary of the road where the vehicle is located can be obtained, and the vertical distance between the vehicle position and the right boundary line of the road can be calculated. The vertical distance is: the vehicle center distance is located in front of the conflict area The distance x from the rightmost boundary line of the road.

Here, it should be noted that if the current meeting state belongs to the parking type, the vehicle needs to drive to the far right of the road to provide sufficient space for the opposite vehicle.

In the present invention, in terms of specific implementation, in step S3, the target speed of the vehicle is determined according to the type of meeting vehicle to which the current meeting state belongs, which specifically includes the following steps:

If the current meeting state is of the speed limit type, the target speed of the vehicle is the preset first target speed (for example, about 20 km/h).

If the current meeting state is an avoidance type, the target speed of the vehicle is the preset second target speed (for example, about 10km/h);

In the present invention, in terms of specific implementation, in step S3, the following steps are also included:

Among them, v _ego represents the current speed of the vehicle;

v _goal represents the target speed of the vehicle (ie the target speed);

v _obj represents the current speed of the opposite vehicle;

y represents the distance between the current position of the vehicle and the opposite vehicle (specifically, the vertical distance between the frontmost position of the front of the vehicle and the frontmost position of the front of the opposite vehicle).

It should be noted that the meeting position is: the position of the vehicle when the longitudinal distance between the opposite vehicle and the vehicle is zero.

It should also be noted that the current speed _vego of the vehicle is provided by a positioning module installed on the vehicle. The vehicle speed v _obj and the relative distance y of the opposite vehicle are provided by the existing perception module installed on the vehicle, and the target speed of the vehicle is obtained by the above logic (for example, about 20km/h, or about 10km/h).

In the present invention, through the above formula, the deceleration distance for the vehicle to reach the meeting position can be calculated and obtained, and the specific position of the meeting position does not need to be obtained here.

It should be noted that, for the present invention, in order to reduce the speed of the vehicle at the meeting moment, it is necessary to estimate the position at the meeting moment (ie the meeting position), and determine the target speed of the meeting position according to the type of the meeting. If the current speed limit type, the vehicle target speed can be about 20km/h, if the current avoidance type, the vehicle target speed can be about 10km/h.

It should also be noted that, for the present invention, if the current meeting type is a parking type, the vehicle needs to stop in front of the conflict area and wait for the opposite vehicle to pass, so the deceleration distance is the position of the vehicle (specifically, the frontmost position of the front of the vehicle) and the conflict. The vertical distance between areas, the target speed is 0km/h.

The position of the conflict area has been described above, the position of the vehicle is obtained through an existing positioning module, and the vertical distance between the vehicle and the conflict area is the detection distance for the presence of the conflict described above.

Based on the above technical solutions, compared with the prior art, the present invention proposes a method for judging and classifying a car meeting scene, and the target end point of the car meeting scene can be planned and converted into lateral offset distance and target speed calculation. At the same time, according to different types of vehicles, the corresponding planning target end point calculation method and key parameter design principles are proposed.

In the present invention, the planned target end point of the vehicle (that is, the end point of the reference path of the vehicle) is determined by the lateral offset distance and the target speed limit value (that is, the target speed). Algorithms such as hybrid A* and quintic polynomial optimization are used to obtain the reference path of the vehicle and the end point of the reference path. This is an existing mature technology and will not be repeated here.

Among them, the calculation method and key parameters of the lateral offset distance and the target speed limit value (ie, the target speed) depend on the type of passing vehicles, as described above.

In addition, based on the above-mentioned method for processing an automatic driving meeting scene provided by the present invention, in order to execute the above-mentioned automatic driving meeting scene processing method, the present invention also provides an automatic driving meeting scene processing device, and the device includes the following modules:

The meeting vehicle type judging processing module is connected with the meeting vehicle state judging module, and is used to classify the current meeting vehicle state after being triggered to run by the meeting vehicle state judging module, judge the corresponding meeting vehicle type, and then according to the preset The corresponding relationship between a plurality of different car meeting types and a plurality of different car meeting processing operations is performed, and the corresponding car meeting processing operations are performed;

In the present invention, based on the above-mentioned method for processing an automatic driving meeting scene provided by the present invention, the automatic driving meeting scene processing device provided by the present invention further includes the following modules:

The driving planning module is connected with the planning target end point determination module, and is used for generating the distance between the vehicle from the current position to the planning target end point according to the current position of the vehicle and the position of the planning target end point of the vehicle determined by the planning target end point determination module. Driving trajectories (eg using existing search algorithms or other existing algorithms capable of generating driving paths).

It should be noted that the driving trajectory planning, which is an existing driving path planning method, can be determined by using an existing well-known method, which will not be repeated here.

In addition, the present invention also provides a vehicle, where the vehicle includes the foregoing apparatus for processing a car meeting scene of an automatic driving.

It should be noted that, in the present invention, in terms of specific implementation, the vehicle deceleration process is estimated when calculating the target speed in the present invention, which can also be based on kinematic models such as trapezoidal acceleration and minimum impact.

In the present invention, in terms of specific implementation, after the planned target end point is calculated in the present invention, sampling can be performed around to generate several planned end points for generating motion trajectories.

Compared with the prior art, the present invention provides a method for processing a car meeting scene of automatic driving, a device, a vehicle, and a storage medium thereof, which have the following beneficial technical effects:

1. The method for processing the meeting scene designed by the present invention can ensure that the automatic driving vehicle maintains a reasonable speed during the meeting, and there is sufficient space for both the vehicle and the opposite vehicle to pass, which improves the driving safety and conforms to human driving. habit, thereby enhancing the passenger's riding experience.

2. The planning end point selection method designed by the present invention fully combines the characteristics of the vehicle driving environment and the current meeting scene type, and is converted into the calculation of the lateral right offset distance and the target speed, and the implementation process is concise and clear.

To sum up, compared with the prior art, the present invention provides a method for processing a car meeting scene of automatic driving, a device, a vehicle, and a storage medium thereof, which are scientifically designed and can actively adjust the vehicle when encountering a car meeting scene. Therefore, the speed of the vehicle at the moment of meeting is reduced, and the lateral separation distance between the vehicle and the opposite vehicle is increased, which is conducive to further improving the safety of driving and satisfying people's driving habits, which has great practice. significance.

The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made. It should be regarded as the protection scope of the present invention.

Claims

A method for processing a car meeting scene of automatic driving, characterized in that it comprises the following steps:

Step S1, determine whether the vehicle is in the meeting state, and when the vehicle is in the meeting state, execute step S2;

Step S2, classifying the current meeting state, judging the corresponding meeting type, according to the preset correspondence between a plurality of different meeting types and a plurality of different meeting processing operations, and executing the current meeting state. The corresponding car meeting processing operation.
The method for processing a car meeting scenario of automatic driving according to claim 1, further comprising the following steps after step S2:

Step S3, according to the type of meeting vehicle to which the current meeting state determined in step S2 belongs, determine the lateral rightward offset distance and target speed of the vehicle;

Step S4, according to the lateral rightward offset distance and the target speed of the vehicle determined in step S3, the planned target end point of the vehicle is obtained, and the planned target end point of the vehicle is the end point of the reference path of the vehicle;

Step S5, according to the current position of the vehicle and the position of the planned target end point of the vehicle determined in step S4, generate a driving track between the vehicle from the current position to the planned target end point.
The method for processing a car meeting scenario of automatic driving according to claim 1 or 2, wherein in step S11, it is judged whether the vehicle meets the preset meeting conditions, if it meets the preset meeting conditions, step S12 is executed, otherwise, it is judged The vehicle is not in a meeting state and will not be processed;

The preset vehicle meeting conditions are specifically: the vehicle is located in the leftmost lane, and there is no physical obstruction on the left side of the vehicle;

Step S12, screen all the opposite vehicles in front of the vehicle, if there is a situation where the opposite vehicle is driving in the adjacent opposite lane of the lane where the vehicle is located, it is judged that the vehicle is in a meeting state at this time, and step S2 is continued, otherwise, it is judged that the vehicle is not. In the state of meeting vehicles, no processing will be performed.
The method for processing a car meeting scene of automatic driving according to any one of claims 1-3, wherein the step S2 specifically includes the following sub-steps:

Step S21, judging whether there is a collision risk between the predicted trajectory of the opposite vehicle and the reference path of the vehicle; if there is no collision risk, it is judged that the current meeting state is of a speed limit type, and the corresponding meeting processing operation is: reduce the speed of the vehicle; If there is a collision risk, it is determined that the current meeting state does not belong to the speed limit type, and step S22 is executed;

Step S22, determine whether there is a conflict area between the vehicle and the opposite vehicle; if there is no conflict area, it is determined that the current vehicle meeting state belongs to the avoidance type, and the corresponding vehicle meeting processing operation is: control the vehicle to perform a lateral right offset motion to avoid Collision; if there is a conflict area, it is determined that the current meeting vehicle state belongs to the parking type, and the corresponding vehicle meeting processing operation is: control the vehicle to stop before entering the conflict area and wait for the opposite vehicle to pass.
The method for processing a car meeting scenario of autonomous driving according to claim 4, wherein in step S21, judging whether there is a collision risk between the predicted trajectory of the opposite vehicle and the reference path of the vehicle, specifically includes:

Calculate the distance between any position point of the predicted trajectory of the opposite vehicle and any position point in the reference path of the vehicle; judge whether there is a distance value smaller than the preset safety distance, if it exists, then judge that there is a collision risk, if not then it is judged that there is no risk of collision;

In the step S22, it is determined whether there is a conflict area between the vehicle and the opposite vehicle, which specifically includes:

When there is a static obstacle in the opposite lane, and the width of the remaining space on the right side of the static obstacle on the opposite lane is less than the sum of the vehicle width, the width of the opposite vehicle and the preset safety distance, then it is judged that the vehicle is in the opposite direction. There is a conflict area between the vehicles.
The method according to claim 2, characterized in that, in step S3, according to the type of the meeting vehicle to which the current meeting state determined in step S2 belongs, the lateral rightward offset distance of the vehicle is determined, and the specific Include the following steps:

If the current meeting state is of the speed limit type, the vehicle's lateral right offset distance is zero;

If the current meeting vehicle state is an avoidance type, calculate the lateral rightward offset distance of the vehicle according to the following calculation formula. The specific formula is as follows:

s=l+0.5w+d;

Wherein, l represents the straight-line distance between the rightmost point of the predicted trajectory of the opposite vehicle and the reference path of the vehicle;

Among them, w represents the width of the vehicle; d represents the preset safety distance, that is, the horizontal vertical distance maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle;

If the current meeting state belongs to the parking type, calculate the lateral rightward offset distance of the vehicle according to the following calculation formula. The specific formula is as follows:

s=x-d;

Among them, x represents the distance from the center of the vehicle to the rightmost boundary line of the road in front of the conflict area;

d represents the preset safety distance, that is, a certain horizontal and vertical distance is maintained between the straight line on the left side of the vehicle and the straight line on the far right side of the opposite vehicle;

In step S3, the target speed of the vehicle is determined according to the type of meeting vehicle to which the current meeting state belongs, which specifically includes the following steps:

If the current meeting state is of the speed limit type, the target speed of the vehicle is the preset first target speed;

If the current meeting state is an avoidance type, the target speed of the vehicle is the preset second target speed;

Wherein, the preset second target speed value is smaller than the preset first target speed value;

If the current meeting state is of the parking type, the target speed of the vehicle is zero.
The method for processing a car meeting scene of automatic driving according to claim 2, wherein, in step S3, it further comprises the following steps:

Control the vehicle to perform a uniform deceleration motion, and calculate the deceleration distance for the vehicle to reach the meeting position. The specific formula is as follows:

Among them, v ego represents the current speed of the vehicle;

v goal represents the target speed of the vehicle;

v obj represents the current speed of the opposite vehicle;

y represents the distance between the current position of the vehicle and the opposite vehicle.
An automatic driving meeting scene processing device, characterized in that it includes the following modules:

The meeting vehicle state judgment module is used to judge whether the vehicle is in the meeting vehicle state. When the vehicle is in the meeting vehicle state, it sends a trigger signal to the meeting vehicle type judgment processing module to trigger the operation of the meeting vehicle type judgment module;

The meeting vehicle type judging processing module is connected with the meeting vehicle state judging module, and is used to classify the current meeting vehicle state after being triggered to run by the meeting vehicle state judging module, and determine the corresponding meeting vehicle type, according to the preset The correspondence between a plurality of different meeting types and a plurality of different meeting processing operations is performed, and a meeting processing operation corresponding to the current meeting state is executed.
automatic driving meeting car scene processing device as claimed in claim 8, is characterized in that, also comprises following module:

The lateral rightward offset distance and target speed determination module is connected with the meeting vehicle type judgment processing module, and is used for determining the lateral direction of the vehicle according to the type of meeting vehicle to which the current meeting vehicle state determined by the meeting vehicle type judgment module belongs. Right offset distance and target speed;

The planning target end point determination module is connected with the lateral right offset distance and the target speed determination module, and is used for determining the lateral right offset distance and the target speed of the vehicle according to the lateral right offset distance and the target speed determination module. , obtain the planned target end point of the vehicle, and the planned target end point of the vehicle is the end point of the reference path of the vehicle;

The driving planning module is connected with the planning target end point determination module, and is used for generating the distance between the vehicle from the current position to the planning target end point according to the current position of the vehicle and the position of the planning target end point of the vehicle determined by the planning target end point determination module. driving track.
A vehicle, characterized by comprising the automatic driving meeting scene processing device according to claim 8 or 9.
A computer-readable storage medium, the computer-readable storage medium stores a program, when the program is executed by a multi-core processor, the multi-core processor is made to execute the multi-core processor according to any one of claims 1-7. method.