Driverless vehicle trafficability prediction and obstacle avoidance cooperative control method

Abstract

The invention relates to the field of driverless car control, and particularly discloses a driverless car trafficability prediction and obstacle avoidance cooperative control method, which comprises the following steps: (1) detecting the height, width and longitudinal distance of an obstacle in front of a vehicle through a camera, a millimeter wave radar and a laser radar; (2) according to the longitudinal distance range of the obstacle and the vehicle, fusing the heights of the obstacle measured by the millimeter wave radar, the laser radar and the camera by adopting a distance-based weighted fusion calculation method to obtain a final height value of the obstacle; (3) based on the center line of the vehicle body as a base line, calculating the transverse distance between the obstacle and the base line, and judging that the obstacle is located outside the vehicle body, under a chassis of the vehicle body or under wheels; and (4) judging whether the vehicle can cross the obstacle or not according to the obstacle position and the obstacle height, and selecting a deceleration crossing or obstacle avoidance strategy according to a judgment result.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned vehicle control, in particular to a kind of unmanned vehicle passability prediction and obstacle avoidance collaborative control method. BACKGROUND

[0002] At present, the safety model of unmanned vehicle mainly distributes in the research of vehicle active collision avoidance and path trajectory planning based on driving risk identification in structured road environment. In such research, the detection and dynamic tracking of target mainly focus on vehicles, pedestrians, bicycles and buildings and other large obstacles, and then emergency braking or steering lane avoidance is taken according to the distance and speed information of the obstacle. Small obstacles such as gravel, vehicle debris and other small obstacles are not concerned. The design of vehicle passability is mainly concentrated in the process of vehicle design, and the design of vehicle body parameters is mainly in the process of research and development. In the process of vehicle driving, the control of vehicle passability is generally achieved by lifting the vehicle body through air suspension to improve the passability of vehicle driving. However, such method has certain defects, ① there is no quantitative measurement of the key information such as height and width of the obstacle, which is easy to cause misjudgment, that is, it is felt that the vehicle body can be lifted to cross the obstacle, but in fact the chassis will collide with the obstacle and be lifted; ② the height of air suspension for lifting the vehicle body is limited, which cannot meet all conditions.

[0003] When the vertical height of small obstacles on the road is greater than the minimum ground clearance of the vehicle, the chassis of the vehicle will collide with the obstacle, or the wheels will run over the small obstacle at high speed, which will cause wheel jumping, side deviation or high intensity vibration, seriously affecting the safety, passability and smoothness of the vehicle. Therefore, the control strategy of unmanned vehicle obstacle avoidance system needs to consider the passability of the vehicle according to the shape information of the road obstacle. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a kind of unmanned vehicle passability prediction and obstacle avoidance collaborative control method, small obstacles are detected, the height, width and position information of the obstacle are judged, then the minimum ground clearance of the vehicle itself is combined to predict whether the front obstacle can pass smoothly, if the obstacle can be crossed, it is directly passed, if the obstacle cannot be crossed, braking or lane changing and other obstacle avoidance measures are taken to solve the problem of vehicle damage caused by small obstacles.

[0005] To achieve the above purpose, the present application is realized by the following technical scheme:

[0006] A kind of unmanned vehicle passability prediction and obstacle avoidance collaborative control method, comprising the following steps:

[0007] (1) the height, width and longitudinal distance of the obstacle in front of the vehicle are detected by camera, millimeter wave radar and laser radar;

[0008] (2) According to the longitudinal distance range of the obstacle and the vehicle, a distance-based weighted fusion calculation method is adopted to fuse the obstacle height measured by the millimeter wave radar, the laser radar and the camera, and an obstacle final height value is obtained;

[0009] (3) Based on the vehicle center line as a baseline, the lateral distance of the obstacle from the baseline is calculated, and it is judged whether the obstacle is located outside the vehicle body, directly below the vehicle chassis or directly below the vehicle wheel;

[0010] (4) If the obstacle is located outside the vehicle body, the vehicle maintains the original driving state; if the obstacle is located directly below the vehicle chassis, the obstacle height is compared with the minimum ground clearance of the vehicle, and when the obstacle height is greater than the minimum ground clearance, the brake or lane change control strategy is triggered; if the obstacle is located directly below the vehicle wheel, the vehicle wheelbase, wheel diameter and road adhesion coefficient are combined to judge whether the vehicle can cross the obstacle, and a deceleration crossing or obstacle avoidance strategy is selected according to the judgment result.

[0011] The present application provides a kind of unmanned vehicle passability prediction and obstacle avoidance coordination control system, comprising:

[0012] (a) multi-sensor module, containing camera, millimeter wave radar and laser radar, for collecting the height, width and position information of obstacle;

[0013] (b) data processing module, for performing distance-based weighted fusion calculation and obstacle position judgment;

[0014] (c) control strategy module, for generating brake, steering or keep driving control instruction according to the comparison result of obstacle position, height and vehicle parameters;

[0015] (d) communication module, control instruction is transmitted to brake system, steering system and vehicle control module through CAN bus.

[0016] The present application further limits the technical scheme:

[0017] Preferably, the distance-based weighted fusion calculation method in step (2) is specifically:

[0018] According to the interval range of the longitudinal distance L of the obstacle and the vehicle, the obstacle height H is calculated by the following weighted formula respectively:

[0019] When L is greater than or equal to 300 meters, H = 0.45H1 + 0.35H2 + 0.2H3;

[0020] When 300 > L ≥ 250 meters, H = 0.4H1 + 0.4H2 + 0.2H3;

[0021] When 250 > L ≥ 200 meters, H = 0.35H1 + 0.45H2 + 0.2H3;

[0022] When 200 > L ≥ 150 meters, H = 0.3H1 + 0.4H2 + 0.3H3;

[0023] When 150 > L ≥ 100 meters, H = 0.25H1 + 0.4H2 + 0.35H3;

[0024] When 100 > L ≥ 50 meters, H = 0.25H1 + 0.35H2 + 0.4H3;

[0025] When 50 > L ≥ 0 meters, H = 0.25H1 + 0.25H2 + 0.5H3;

[0026] Wherein, H1 is the height measured by millimeter wave radar, H2 is the height measured by laser radar, and H3 is the height measured by camera.

[0027] Preferably, the calculation formula of the lateral distance of the obstacle in step (3) is:

[0028] Let the straight-line distance between the obstacle and the camera be L, the camera center line (also the vehicle center line) be the baseline, z be the connecting line between the obstacle and the camera, the angle between z and the baseline be θ, the vertical distance between the obstacle and the baseline be m, the width of the tire be K, the wheelbase be B, and the lane width be S. The lateral distance m is calculated through geometric relationship as:

[0029] The value of m is the distance between the obstacle and the vehicle center line, through which the lateral position of the obstacle relative to the vehicle can be represented, and L represents the longitudinal distance of the obstacle from the front of the vehicle (camera).

[0030] Preferably, when the lane changing control strategy is triggered in step (4), the following safety conditions need to be met:

[0031] The safety distance model is:

[0032]

[0033] Wherein, ds is the safety distance; v1 is the vehicle speed; v2 is the speed of the obstacle; if it is a stationary obstacle, v2 = 0; a1 is the vehicle deceleration; a2 is the deceleration of the obstacle; t1 is the deceleration time; t2 is the system delay time, which depends on the system software and hardware; d0 is the distance between the target obstacle and the vehicle after braking, usually 2.5-3 meters;

[0034] Distance d in front of adjacent lane next-f There is no obstacle in the inner lane, and d next-rIf there is no obstacle behind the vehicle with a speed higher than the vehicle speed, the lane changing condition is met, and the following equation is used,

[0035]

[0036] where v r is the speed of the obstacle behind the adjacent lane, v f is the speed of the obstacle in front of the adjacent lane, a f is the acceleration thereof, L is the longitudinal displacement during lane changing, and d s is the minimum braking distance of the vehicle.

[0037] Preferably, the specific method for determining whether the vehicle can cross the obstacle in step (4) is as follows: according to the wheelbase B, wheel diameter D, distance L9 between the front axle and the center of gravity, and road adhesion coefficient μ, the maximum step height P that can be crossed by the vehicle is calculated, if the height H of the obstacle is less than or equal to P, the vehicle can pass through by deceleration; if H>P, the lane changing or braking strategy is triggered.

[0038] Preferably, the data processing module further comprises:

[0039] an obstacle height fusion unit configured to dynamically adjust the weight coefficients of the millimeter wave radar, laser radar and camera according to the longitudinal distance interval;

[0040] an obstacle position classification unit configured to divide the obstacle position into three types, i.e., outside the vehicle body, directly below the chassis, or directly below the wheel.

[0041] Preferably, the control strategy module further comprises:

[0042] a crossing ability evaluation unit configured to calculate the maximum height of the obstacle that can be crossed by the vehicle in combination with the wheelbase, wheel diameter and road adhesion coefficient;

[0043] a safety distance calculation unit configured to dynamically update the minimum safety distance required for lane changing or braking.

[0044] Advantages

[0045] Compared with the prior art, the present application has the following advantages:

[0046] 1. The present application uses three types of sensors, i.e., camera, millimeter wave radar and laser radar, which are combined and can improve the detection accuracy of the height and width of small obstacles in bad weather through a distance-based weighting measurement method.

[0047] 2、The obstacle position judgment method according to the vehicle body center line as a reference, and the obstacle position is classified into 3 categories, i.e., located outside the vehicle body, located directly below the vehicle body and located directly below the wheel, which lays a foundation for the passability control of the unmanned vehicle.

[0048] 3、The passability pre-judgment method; according to the different positions of the obstacles, the size of the obstacle height and width, different control methods are selected, so that the unmanned vehicle can also maintain good passability and obstacle avoidance performance in the occasional rock and other occasional convex road surface state. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 The method flowchart of the embodiment is provided. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0051] Embodiment 1

[0052] The embodiment provides an unmanned vehicle passability prediction and obstacle avoidance collaborative control system, which comprises:

[0053] (a) a multi-sensor module comprising a camera, a millimeter wave radar and a laser radar, for collecting height, width and position information of obstacles;

[0054] (b) a data processing module for performing distance-based weighted fusion calculation and obstacle position judgment, which comprises:

[0055] an obstacle height fusion unit for dynamically adjusting the weight coefficients of the millimeter wave radar, the laser radar and the camera according to the longitudinal distance interval;

[0056] an obstacle position classification unit for classifying the obstacle position into three types of outside the vehicle body, directly below the chassis or directly below the wheel;

[0057] (c) a control strategy module for generating control instructions of braking, steering or keeping driving according to the comparison results of the obstacle position, height and vehicle parameters, which comprises:

[0058] a crossing ability evaluation unit for calculating the limit height of the vehicle that can cross the obstacle in combination with the wheelbase, wheel diameter and road adhesion coefficient of the vehicle;

[0059] A safety distance calculation unit is configured to dynamically update a minimum safety distance required for lane changing or braking.

[0060] (d) a communication module configured to transmit control instructions to a braking system, a steering system, and a vehicle control module via a CAN bus.

[0061] Based on the above system, the embodiment provides a self-driving vehicle passability prediction and obstacle avoidance collaborative control method, comprising the following steps:

[0062] (1) detecting the height, width and longitudinal distance of the obstacle in front of the vehicle through a camera, a millimeter wave radar and a laser radar;

[0063] (2) according to the longitudinal distance range of the obstacle and the vehicle, a distance-based weighted fusion calculation method is used to fuse the obstacle height measured by the millimeter wave radar, the laser radar and the camera, and obtain the final obstacle height value, wherein the distance-based weighted fusion calculation method is specifically:

[0064] According to the interval range of the longitudinal distance L of the obstacle and the vehicle, the following weighted formula is used to calculate the obstacle height H:

[0065] When L is greater than or equal to 300 meters, H = 0.45H1 + 0.35H2 + 0.2H3;

[0066] When 300 > L is greater than or equal to 250 meters, H = 0.4H1 + 0.4H2 + 0.2H3;

[0067] When 250 > L is greater than or equal to 200 meters, H = 0.35H1 + 0.45H2 + 0.2H3;

[0068] When 200 > L is greater than or equal to 150 meters, H = 0.3H1 + 0.4H2 + 0.3H3;

[0069] When 150 > L is greater than or equal to 100 meters, H = 0.25H1 + 0.4H2 + 0.35H3;

[0070] When 100 > L is greater than or equal to 50 meters, H = 0.25H1 + 0.35H2 + 0.4H3;

[0071] When 50 > L is greater than or equal to 0 meters, H = 0.25H1 + 0.25H2 + 0.5H3;

[0072] Wherein, H1 is the height measured by the millimeter wave radar, H2 is the height measured by the laser radar, and H3 is the height measured by the camera;

[0073] (3) based on the center line of the vehicle body as a baseline, the lateral distance of the obstacle from the baseline is calculated to determine whether the obstacle is located outside the vehicle body, directly below the vehicle chassis or directly below the wheels, wherein the calculation formula of the lateral distance of the obstacle is:

[0074] Let the straight-line distance between the obstacle and the camera be L, the camera center line (also the vehicle center line) be the base line, z be the connecting line between the obstacle and the camera, the angle between z and the base line be θ, the vertical distance between the obstacle and the base line be m, the width of the tire be K, the wheelbase be B, and the lane width be S. The lateral distance m is calculated by geometric relationship as follows:

[0075] m = L sin θ

[0076] The value of m is the distance between the obstacle and the vehicle center line, by which the lateral position of the obstacle relative to the vehicle can be represented, and L represents the longitudinal distance of the obstacle from the front of the vehicle (the camera).

[0077] (4) If the obstacle is located outside the vehicle body, the vehicle maintains the original driving state. If the obstacle is located directly below the vehicle chassis, the height of the obstacle is compared with the minimum ground clearance of the vehicle. When the height of the obstacle is greater than the minimum ground clearance, the brake or lane change control strategy is triggered. If the obstacle is located directly below the wheel, the vehicle wheelbase, wheel diameter, and road adhesion coefficient are combined to determine whether the vehicle can cross the obstacle, and the deceleration crossing or obstacle avoidance strategy is selected according to the determination result, which is specifically as follows:

[0078] When the autonomous vehicle control system detects the lateral position m of the obstacle, the system performs data comparison processing, and the processing method is as follows:

[0079] When the system compares and obtains: or

[0080] It indicates that the obstacle is not in front of the vehicle chassis or the wheel, and no processing is required. The autonomous vehicle continues to work according to the original set route, speed, and driving action. At this time, the control strategy can be represented by the following formula:

[0081]

[0082] Con represents the control strategy of the passability module (computer) of the autonomous vehicle; f ecu represents the control strategy of the vehicle control module (computer) of the autonomous vehicle; the passability control module and the vehicle control module communicate through the CAN bus.

[0083] When the system determines that at this time, it indicates that the obstacle is located below the chassis, and the minimum ground clearance and the jack failure need to be considered.

[0084] At this time, the system needs to further compare the size between the height of the obstacle and the minimum ground clearance:

[0085] When h > H, it means that the height of the obstacle is less than the minimum ground clearance, at this time the vehicle directly passes through the obstacle without scratching, and the autonomous vehicle continues to work according to the original route, vehicle speed and driving action set by the vehicle control module;

[0086]

[0087] When h≤H, it means that the height of the obstacle is greater than and equal to the minimum ground clearance, at this time the collision and lifting failure will occur, and the vehicle cannot directly pass through the obstacle, at this time the autonomous vehicle needs to use further control strategy, and then the rear camera and millimeter wave scan the traffic conditions of the adjacent lane behind the vehicle;

[0088] When or , it means that the obstacle is located directly below the wheel and tire, at this time the ability of the autonomous vehicle to cross the trench and step needs to be considered;

[0089] According to the known vehicle wheelbase B, wheel diameter D, front axle distance from the center of gravity L9 and road adhesion coefficient μ, the maximum step height P that the vehicle can cross is calculated, if the obstacle height H≤P, the vehicle slows down to pass; if H>P, the lane changing or braking strategy is triggered, and the following control needs to be performed:

[0090] When P≥H, it means that the vehicle can cross the obstacle, at this time the vehicle slows down to cross the obstacle. The vehicle speed control method when crossing the obstacle is as follows:

[0091]

[0092] con v is the vehicle speed controlled by the controller.

[0093] When P<H, it means that the vehicle cannot cross the obstacle, at this time the vehicle needs to perform lane changing or braking action.

[0094] When it is judged that the obstacle cannot be passed, the control strategy of vehicle obstacle avoidance, braking and steering compatibility needs to be considered, wherein the following equation represents the safety distance model:

[0095]

[0096] Where d s is the safety distance; v1 is the vehicle speed; v2 is the speed of the obstacle; if it is a static obstacle, v2=0; a1 is the vehicle deceleration; a2 is the deceleration of the obstacle; t1 is the deceleration time; t2 is the system delay time, which depends on the system software and hardware; d0 is the distance between the target obstacle and the vehicle after braking, usually 2.5-3 meters;

[0097] When the target obstacle is in a static state or the distance between the vehicle and the dynamic obstacle in front is less than the safe distance, the lane changing strategy should be given priority, at this time, it is necessary to consider whether the adjacent lane meets the lane changing requirements. If there is no obstacle in front of the adjacent lane within a given range d next-f , and there is no obstacle higher than the vehicle speed behind the set range d next-r , the lane changing condition is met, which is shown in the following equation,

[0098]

[0099] where v r is the speed of the obstacle behind the adjacent lane, v f is the speed of the obstacle in front of the adjacent lane, a f is its acceleration, and L is the longitudinal displacement during lane changing. d s is the minimum braking distance of the vehicle, and the vehicle stops by braking when neither of the left and right lanes meets the lane changing condition.

[0100] At the beginning of this embodiment, the system uses millimeter wave radar, laser radar and camera to scan whether there is a small obstacle in front of the vehicle, respectively. After detecting the obstacle, the height and width of the obstacle are measured, respectively. Then, according to the working characteristics of millimeter wave radar, laser radar and camera, a distance-based weighted measurement method is used to fuse the height and width information, and the obstacle height and width information with higher reliability is obtained. Then, the longitudinal distance between the obstacle and the vehicle body is measured, and finally the lateral distance between the obstacle and the vehicle center line is detected, and the more complete obstacle position information is obtained.

[0101] The position relationship between the obstacle and the vehicle body is judged, specifically:

[0102] ① If the obstacle is at the edge of the lane and not directly below the chassis or tire, the autonomous vehicle continues to move forward according to the original speed and route set by the driver;

[0103] ② When the obstacle is directly below the chassis, the relationship between the height of the obstacle and the minimum ground clearance of the vehicle is judged. If the height of the obstacle is less than the minimum ground clearance, the autonomous vehicle continues to move forward according to the original speed and route set by the driver. If the height of the obstacle is greater than the minimum ground clearance, the vehicle needs to be stopped or changed lane. At this time, the vehicle information of the adjacent lane needs to be detected. If the adjacent lane behind the vehicle meets formula (a), the lane changing process is taken; if there is no vehicle in the adjacent lane behind the vehicle, the speed reduction and stopping strategy is taken.

[0104] If the obstacle is directly below the tire of the vehicle, it is considered whether the vehicle can cross the obstacle. The value between the height of the obstacle and the limit height that the vehicle can cross is determined. If the height of the obstacle is greater than the limit height that the vehicle can cross, the vehicle information of the adjacent lane is detected. If the rear adjacent lane satisfies formula (a), indicating that there is no vehicle within 300 meters in the adjacent lane, the lane changing process is taken. If the rear adjacent lane does not satisfy formula (a), indicating that there is a vehicle in the adjacent lane, the vehicle cannot change lanes, and the deceleration and parking strategy is taken. If the height of the obstacle is less than the limit height that the vehicle can cross, it indicates that the vehicle can pass through the obstacle. Considering the safety and comfort during the crossing process, the vehicle is driven according to the pre-set control strategy between the vehicle speed and the height of the obstacle.

[0105] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that are not specifically enumerated herein. The use of the terms "including," "comprising," or "having" and variations thereof herein is intended to be equivalent to the term "comprising" so as to encompass the non-exclusive inclusion of the listed elements and the possibility that other elements not expressly listed are also included. Moreover, the use of the term "comprising" or "including" or "having" is not intended to limit the present application to only those embodiments which contain the recited elements, but rather to allow for other embodiments to be included which do not contain the recited elements.

[0106] While the embodiments of the application have been shown and described, it is to be understood that the application is not limited to the details of the embodiments described, since the scope of the expenditure of the application is defined with respect to the appended claims.

Claims

1. A method for cooperative control of autonomous vehicle passability prediction and obstacle avoidance, characterized in that, Includes the following steps: (1) Detect the height, width and longitudinal distance of obstacles in front of the vehicle using cameras, millimeter-wave radar and lidar; (2) Based on the longitudinal distance range between the obstacle and the vehicle, a distance-based weighted fusion calculation method is used to fuse the obstacle height measured by the millimeter-wave radar, lidar and camera to obtain the final obstacle height value; (3) Based on the center line of the vehicle body as the baseline, calculate the lateral distance between the obstacle and the baseline, and determine whether the obstacle is located outside the vehicle body, directly below the chassis of the vehicle body, or directly below the wheel; (4) If the obstacle is located outside the vehicle body, the vehicle maintains its original driving state; if the obstacle is located directly below the vehicle chassis, the height of the obstacle is compared with the minimum ground clearance of the vehicle. When the height of the obstacle is greater than the minimum ground clearance, the braking or lane change control strategy is triggered. If the obstacle is located directly under the wheels, the vehicle's wheelbase, wheel diameter, and road surface adhesion coefficient are considered to determine whether the vehicle can cross the obstacle, and a strategy of decelerating to cross or avoiding the obstacle is selected based on the judgment result.

2. The autonomous vehicle passability prediction and obstacle avoidance cooperative control method according to claim 1, characterized in that, The distance-based weighted fusion calculation method in step (2) is as follows: Based on the range of the longitudinal distance L between the obstacle and the vehicle, the obstacle height H is calculated using the following weighted formula: When L ≥ 300 meters, H = 0.45H1 + 0.35H2 + 0.2H3; When 300 > L ≥ 250 meters, H = 0.4H1 + 0.4H2 + 0.2H3; When 250 > L ≥ 200 meters, H = 0.35H1 + 0.45H2 + 0.2H3; When 200 > L ≥ 150 meters, H = 0.3H1 + 0.4H2 + 0.3H3; When 150 > L ≥ 100 meters, H = 0.25H1 + 0.4H2 + 0.35H3; When 100 > L ≥ 50 meters, H = 0.25H1 + 0.35H2 + 0.4H3; When 50 > L ≥ 0 meters, H = 0.25H1 + 0.25H2 + 0.5H3; Where H1 is the height measured by millimeter-wave radar, H2 is the height measured by lidar, and H3 is the height measured by camera.

3. The autonomous vehicle passability prediction and obstacle avoidance cooperative control method according to claim 1, characterized in that, The formula for calculating the lateral distance of the obstacle in step (3) is as follows: Let L be the straight-line distance between the obstacle and the camera, the camera centerline (which is also the vehicle centerline) be the baseline, z be the line connecting the obstacle and the camera, the angle between z and the baseline be θ, the vertical distance between the obstacle and the baseline be m, the tire width be K, the wheelbase be B, and the lane width be S. Then the lateral distance m is calculated using geometric relationships as follows: m=L·sinθ The value of m is the distance between the obstacle and the centerline of the vehicle. It can represent the lateral position of the obstacle relative to the vehicle. L represents the longitudinal distance between the obstacle and the frontmost part of the vehicle (camera).

4. The autonomous vehicle passability prediction and obstacle avoidance cooperative control method according to claim 1, characterized in that, When the lane change control strategy is triggered in step (4), the following safety conditions must be met: The safe distance model is as follows: Where ds is the safety distance; v1 is the vehicle speed; v2 is the obstacle speed; if the obstacle is stationary, v2 = 0; a1 is the vehicle deceleration; a2 is the obstacle deceleration; t1 is the deceleration time; t2 is the system delay time, which depends on the system software and hardware; d0 is the distance between the target obstacle and the vehicle after braking, usually taken as 2.5-3 meters; Distance d ahead of adjacent lane next-f There are no obstacles inside, and d next-r If there are no obstacles behind the vehicle that exceed its speed, then the lane change conditions are met, as shown in the following equation. Where v r It is the speed of the obstacle behind the adjacent lane, v f It is the rotational speed of the obstacle ahead of the adjacent lane, a f It is its acceleration, L is the longitudinal displacement during the lane change, and d is the acceleration. s It is the minimum braking distance of a vehicle. When neither the left nor right lane of the vehicle meets the conditions for changing lanes, the vehicle will brake to a stop.

5. The autonomous vehicle passability prediction and obstacle avoidance cooperative control method according to claim 1, characterized in that, The specific method for determining whether a vehicle can cross an obstacle in step (4) is as follows: Based on the vehicle's wheelbase B, wheel diameter D, distance between the front axle and the center of gravity L9, and road surface adhesion coefficient μ, calculate the maximum step height P that the vehicle can cross. If the obstacle height H≤P, the vehicle decelerates and passes through; if H>P, a lane change or braking strategy is triggered.

6. A collaborative control system for predicting the passability and avoiding obstacles of an unmanned vehicle, characterized in that, include: (a) A multi-sensor module, including a camera, millimeter-wave radar and lidar, is used to collect information on the height, width and location of obstacles; (b) Data processing module, used to perform distance-based weighted fusion calculations and obstacle location determination; (c) Control strategy module, used to generate control commands for braking, steering or maintaining driving based on the comparison results of obstacle position and height with vehicle parameters; (d) Communication module, which transmits control commands to the braking system, steering system and vehicle control module via CAN bus.

7. The autonomous vehicle passability prediction and obstacle avoidance cooperative control system according to claim 6, characterized in that, The data processing module further includes: An obstacle fusion unit is used to dynamically adjust the weighting coefficients of millimeter-wave radar, lidar, and camera according to the longitudinal distance range; The obstacle location classification unit is used to classify obstacle locations into three types: outside the vehicle body, directly under the chassis, or directly under the wheels.

8. The autonomous vehicle passability prediction and obstacle avoidance cooperative control system according to claim 6, characterized in that, The control strategy module further includes: The obstacle crossing capability assessment unit is used to calculate the maximum height that a vehicle can cross over obstacles by combining the vehicle's wheelbase, wheel diameter, and road surface adhesion coefficient. The safety distance calculation unit is used to dynamically update the minimum safe distance required for lane changing or braking.

Images