A forward following platooning system and method for a fleet of vehicles

Abstract

The application relates to a forward-following linkage driving system and method for a vehicle fleet, which comprises a leading vehicle and a plurality of unmanned following vehicles following the driving of the leading vehicle. The unmanned following vehicle is provided with a look-around danger early warning module for acquiring environmental information around the current vehicle; a longitudinal dynamic distance adjustment module for monitoring the distance between the current vehicle and the leading vehicle in real time and calculating the heading angle of the leading vehicle during driving; a lateral lane-changing obstacle avoidance module for detecting existing traffic obstacles on the driving track of the leading vehicle in real time; a vehicle-vehicle communication module for establishing communication between the unmanned following vehicle and the leading vehicle; a central control module for receiving information of the above modules, analyzing and deciding according to the information, and sending decision information to a vehicle actuator; and the vehicle actuator for receiving and executing the decision information of the central control module. Compared with the prior art, the application can guarantee the stability, safety and reliability of the unmanned driving vehicle fleet.

Description

Technical Field

[0001] This invention relates to the field of convoy formation technology for intelligent driving, and in particular to a forward following linkage driving system and method for convoy formation. Background Technology

[0002] With the rapid development of technologies such as cameras, millimeter-wave radar, LiDAR, deep learning, and 5G networks, research on autonomous driving technology has made leaps and bounds. However, in real-world traffic conditions, factors such as complex road conditions, numerous traffic participants, and limited information exchange between vehicles make it difficult to achieve unmanned driving under all conditions. In recent years, with the rise of the platooning concept, the method of using a lead vehicle to guide following vehicles in unmanned driving has solved many autonomous driving problems and promoted the development of the field.

[0003] However, in existing autonomous driving technology solutions for platoon formation, unmanned following vehicles simply rely on the driving trajectory of the vehicle in front to travel on the same line, ignoring the dynamics of traffic conditions and lacking active safety driving technologies such as autonomous perception, longitudinal dynamic adjustment of vehicle distance, and lateral obstacle avoidance, which cannot effectively guarantee the safety of unmanned platoon formation. Summary of the Invention

[0004] The purpose of this invention is to overcome the defects of the prior art and provide a forward following linkage driving system and method for convoy formation, which can ensure the stability, safety and reliability of unmanned convoys.

[0005] The objective of this invention can be achieved through the following technical solutions: According to a first aspect of the present invention, the present invention provides a forward-following linkage driving system for convoy formation, comprising a lead vehicle and a plurality of unmanned following vehicles following the lead vehicle, wherein the unmanned following vehicles are equipped with: Surround view hazard warning module: used to obtain environmental information around the current vehicle, determine the distance between the current vehicle and the traffic participants around the current vehicle based on the environmental information, and determine whether to send a warning information to the central control module based on the distance; Longitudinal dynamic distance adjustment module: used to measure the distance between the current vehicle and the guide vehicle in real time, calculate the heading angle of the guide vehicle during the driving process, and send the distance and heading angle information to the central control module; Lateral lane-changing obstacle avoidance module: used to detect existing traffic obstacles on the guide vehicle's trajectory in real time and send the traffic obstacle detection information to the central control module; Vehicle-to-vehicle communication module: used to establish communication between the unmanned following vehicle and the guide vehicle, and to exchange information on the driving status data between the guide vehicle and the unmanned following vehicle; Central control module: Used to receive and analyze warning information, the current distance between the vehicle and the guide vehicle, the heading angle of the guide vehicle, traffic obstacle information and the driving status data of the guide vehicle, and send the decision information to the vehicle actuators; Vehicle actuators: used to receive and execute decision information from the central control module.

[0006] Preferably, the surround-view hazard warning module includes: Surround view monitoring unit: used to monitor the environmental information around the current vehicle; Target detection unit: Used to detect targets based on environmental information, and to detect traffic participants around the current vehicle; Danger warning unit: Used to determine whether the distance between the current vehicle and surrounding traffic participants is less than the safe distance, and if so, to send a warning message to the central control module.

[0007] Preferably, the surround view monitoring unit includes multiple fisheye cameras installed on the current vehicle. The spherical images acquired by the multiple fisheye cameras are transformed into a square field of view through projection, thereby obtaining the environmental information of the current vehicle.

[0008] Preferably, the target detection unit uses Yolov3 as the backbone network, ResNet16 as the encoder, and standard classification cross-entropy and binary entropy loss functions as loss functions to construct and train the target detection model. The environmental information is input into the target detection model to detect traffic participants around the current vehicle.

[0009] Preferably, the longitudinal dynamic distance adjustment module includes: Distance measurement unit: Used to monitor the distance between the current vehicle and all traffic participants within a set range ahead. L ai The system uses an object detection network to detect the guide vehicle among traffic participants, extracts the guide vehicle's position information in front of the current vehicle, and then extracts the distance between the current vehicle and the guide vehicle. L i ; Heading angle calculation unit: used to calculate the heading angle of the guide vehicle. The heading angle is calculated based on the maximum and minimum distances between the current vehicle and the guide vehicle, as well as the lateral distance between the maximum and minimum points.

[0010] Preferably, the formula describing the heading angle of the guide vehicle is as follows: In the formula, and These are the distances between the currently monitored vehicle and the guide vehicle. L iThe maximum and minimum values, This represents the horizontal distance between the maximum and minimum points.

[0011] Preferably, the distance measurement unit obtains the distance between the lead vehicle and the unmanned following vehicle in the convoy formation by data fusion of hybrid solid-state radar and high-definition camera. The hybrid solid-state radar and the high-definition camera are mounted side by side on a bracket, which is located at the center of the front of the current vehicle. The hybrid solid-state radar is used to collect the three-dimensional spatial point cloud information of the vehicle, and the high-definition camera is used to obtain the pixel coordinates of the lead vehicle in the two-dimensional image.

[0012] Preferably, the lateral lane-changing obstacle avoidance module detects in real time whether there are traffic obstacles on the driving trajectory of the guide vehicle. When there are traffic obstacles, it determines whether there are traffic participants within a set range in the left lane of the current vehicle. When there are no traffic participants, it sends information to the central control module, which controls the vehicle actuator to drive the vehicle to change lanes to the left lane and perform an overtaking action. After completing the overtaking maneuver, the lateral lane-changing obstacle avoidance module detects in real time whether there are any traffic participants in the right front lane of the current vehicle. If there are no traffic participants, the central control module controls the vehicle actuator to drive the vehicle back to the movement trajectory of the guide vehicle and continue to follow the guide vehicle for unmanned driving.

[0013] According to a second aspect of the present invention, the present invention provides a method for implementing a forward following linkage driving system for convoy formation as described above, comprising the following steps: S1: Real-time monitoring of the environmental information around the current unmanned following vehicle and the distance between the current unmanned following vehicle and the guide vehicle; real-time detection of existing traffic obstacles on the guide vehicle's driving trajectory; establishment of communication between the current unmanned following vehicle and the guide vehicle; and exchange of information on the driving status data between the guide vehicle and the unmanned following vehicle. S2: Determine the distance between the current unmanned following vehicle and the traffic participants around it based on the environmental information monitored by S1, and decide whether to send a warning message based on the distance; Based on the maximum and minimum distances between the current unmanned following vehicle and the guide vehicle monitored by S1, as well as the lateral distances between the maximum and minimum points, the heading angle of the guide vehicle during the driving process is calculated. S3: Based on the driving status data of the guide vehicle in S1, the distance between the current unmanned following vehicle and the guide vehicle, and traffic obstacle information, as well as the warning information in S2 and the heading angle of the guide vehicle during driving, adjust the driving status of the current unmanned following vehicle, and generate driving status data of the unmanned following vehicle to feed back to the guide vehicle, so as to realize the linkage control between the guide vehicle and the unmanned following vehicle.

[0014] Preferably, the driving status data of the guide vehicle includes throttle opening, steering wheel angle, brake pedal opening angle, and driving trajectory.

[0015] Compared with the prior art, the present invention has the following advantages: This invention utilizes a surround-view hazard warning module to monitor the surrounding environment of the convoy in real time and dynamically adjust the movement status of the unmanned following vehicles, thereby improving the safety performance of the unmanned convoy. A longitudinal dynamic distance adjustment module and a lateral lane-changing obstacle avoidance module acquire real-time information on the guide vehicle's trajectory, heading angle, vehicle-to-vehicle spacing, and traffic participants in front of the unmanned following vehicles in the convoy, and simultaneously adjust the movement status of the unmanned following vehicles, improving the reliability and stability of their forward movement. A V2V vehicle-to-vehicle communication system facilitates wireless communication within the convoy, enabling the convoy to adjust the movement status of the unmanned following vehicles in real time according to traffic conditions, promoting the stability, safety, and integrity of the unmanned forward movement of the convoy. A central control module aggregates real-time data from various sensors within the convoy and dynamically adjusts the movement of the unmanned following vehicles through a closed-loop structure. After integration, all systems and equipment form a complete convoy forward following system, ensuring the stability, safety, and reliability of the unmanned convoy. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a forward-following linkage driving system for platoon formation provided in this embodiment.

[0017] Figure 2 for Figure 1 A schematic diagram of the fisheye camera installation location in the illustrated embodiment.

[0018] Figure 3 for Figure 1 The three-dimensional spatial coordinate system established by the surround view hazard warning module in the illustrated embodiment.

[0019] Figure 4 for Figure 1 A schematic diagram of the surround view field of view constructed by the fisheye camera of the surround view hazard warning module in the embodiment shown.

[0020] Figure 5 for Figure 1 The illustrated embodiment shows a schematic diagram of the installation of the hybrid solid-state radar and high-definition camera.

[0021] Figure 6 for Figure 1 O in the illustrated embodiment 1 X 1 Y 1 Z 1 Three-dimensional spatial coordinate system.

[0022] Figure 7 for Figure 1 The diagram shows a longitudinal dynamic distance adjustment module measuring the heading angle in the embodiment shown.

[0023] Figure 8 for Figure 1 The illustrated embodiment shows the operation diagrams of the lateral lane-changing obstacle avoidance module, including lane changing, overtaking, and trajectory regression.

[0024] Explanation of markings in the diagram: 1. Fisheye camera, 2. High-definition camera, 3. Hybrid solid-state radar, 4. Bracket. Detailed Implementation

[0025] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0026] Example refer to Figure 1 As shown, according to a first aspect of the present invention, this embodiment provides a forward-following linkage driving system for convoy formation, including a lead vehicle and multiple unmanned following vehicles that follow the lead vehicle. The unmanned following vehicles are equipped with: Surround view hazard warning module: used to obtain environmental information around the current vehicle, determine the distance between the current vehicle and the traffic participants around the current vehicle based on the environmental information, and determine whether to send a warning information to the central control module based on the distance; Longitudinal dynamic distance adjustment module: used to measure the distance between the current vehicle and the guide vehicle in real time, calculate the heading angle of the guide vehicle during the driving process, and send the distance and heading angle to the central control module; Lateral lane-changing obstacle avoidance module: used to detect existing traffic obstacles on the guide vehicle's trajectory in real time and send the traffic obstacle information to the central control module; Vehicle-to-vehicle communication module: used to establish communication between the unmanned following vehicle and the guide vehicle, and to exchange information on the driving status data between the guide vehicle and the unmanned following vehicle; Central control module: Used to receive and analyze warning information, the current distance between the vehicle and the guide vehicle, the heading angle of the guide vehicle, traffic obstacle information and the driving status data of the guide vehicle, and send the decision information to the vehicle actuators; Vehicle actuators: used to receive decision information from the central control module and execute it.

[0027] The surround view hazard warning module includes: The surround view monitoring unit is used to monitor the environmental information around the vehicle. The target detection unit is used to detect targets based on environmental information, and to detect traffic participants around the current vehicle; The hazard warning unit is used to determine whether the distance between the current vehicle and surrounding traffic participants is less than the safe distance. If so, it sends a warning message to the central control module.

[0028] refer to Figure 2 As shown, the surround view monitoring unit includes multiple fisheye cameras 1.

[0029] As an optional implementation, the fisheye camera 1 is a 180° fisheye camera.

[0030] As an optional implementation method, refer to Figure 3 As shown, a three-dimensional XYZ coordinate system is established with the vehicle center as the origin, the front of the vehicle as the positive X-axis, the left side of the vehicle as the positive Y-axis, and the top of the vehicle as the positive Z-axis.

[0031] As an optional implementation, four fisheye cameras 1 are used, each installed at an angle of -15° to the positive Z-axis. This installation angle setting enables the establishment of a wide-field-of-view, fully stitched surround-view warning area.

[0032] Specifically, each fisheye camera 1 acquires a spherical image, and the four independent spherical images, which are paired but have overlapping fields of view, are then processed... Figure 4 The projection transformation shown creates a complete square field of view. The gray area in the diagram represents the panoramic field of view, while the area formed by adding length and width is the dynamically adjustable area. This dynamically adjustable area can be adjusted according to the field of view requirements, thereby adjusting the panoramic field of view range. The panoramic monitoring unit can observe environmental information from 0 to 5 meters around the vehicle.

[0033] In existing technologies, object detection networks are built and trained to detect objects in images. However, these networks are only applicable to images captured by high-definition cameras and ignore distortion caused by fisheye cameras. As a result, they cannot accurately detect objects, i.e., they cannot accurately detect traffic participants around vehicles.

[0034] The target detection unit in this embodiment modifies the Darknet53 encoder to ResNet16 based on the Yolov3 network and replaces the L2 loss function with the standard classification cross-entropy and binary entropy loss functions, resulting in target detection that is more suitable for fisheye camera distortion scenarios. It has high performance and can detect traffic participants around the current vehicle in real time.

[0035] Because unmanned vehicles in a convoy are subject to interference from various traffic participants during operation, the lateral distance between them and vehicles and pedestrians is less than the safe distance. The surround-view hazard warning module can help unmanned vehicles maintain a safe distance from other traffic participants over a long period of time, reducing the probability of accidents.

[0036] The longitudinal dynamic distance adjustment module includes: Distance measurement unit: Used to monitor the distance between the current vehicle and all traffic participants within a set range ahead.L ai The system uses an object detection network to detect the guide vehicle among traffic participants, extracts the guide vehicle's position information in front of the current vehicle, and then extracts the distance between the current vehicle and the guide vehicle. L i ; As an optional implementation, the distance between the lead vehicle and the unmanned following vehicle in the convoy formation is obtained by data fusion of hybrid solid-state radar 3 and high-definition camera 2. L i and the heading angle of the guide vehicle ; refer to Figure 5 As shown, the high-definition camera 2 and the hybrid solid-state radar 3 are mounted side by side on the bracket 4, which is located at the center of the front of the unmanned following vehicle.

[0037] In this embodiment, the center of the vehicle's front is taken as the origin, the front of the vehicle is the positive X-axis, the left side of the vehicle is the positive Y-axis, and the top of the vehicle is the positive Z-axis, as shown below. Figure 6 The O shown 1 X 1 Y 1 Z 1 Three-dimensional spatial coordinate system. Support 4 is set at O. 1 Point.

[0038] The distance between the current vehicle and all traffic participants within a 100-meter range is obtained using the hybrid solid-state radar 3 (automotive-grade "Raysun hybrid solid-state radar CH128", ranging range 0-200m). L ai However, it is not possible to accurately extract the distance between the unmanned following vehicle and the lead vehicle at this time; therefore, a high-definition camera 2 is introduced, and the YOLOv3 network is used to extract the position information of the lead vehicle in front of the current vehicle in the convoy formation.

[0039] The hybrid solid-state radar 3 collects three-dimensional spatial point cloud information, with different colors in the point cloud representing different distance values. The high-definition camera 2 acquires the pixel coordinates of the guide vehicle in the two-dimensional image. In the longitudinal dynamic distance adjustment module, the hybrid solid-state radar and the high-definition camera are first jointly calibrated, and then the three-dimensional point cloud coordinates (X,Y,Z) obtained by the hybrid solid-state radar 3 are projected onto the two-dimensional plane of the image using equations (1) and (2) to accurately extract the distance between the unmanned following vehicle and the leading vehicle. .

[0040] (1) In the formula, and O 1 X 1 Y 1 Z 1Scale factors in the X and Y axes of a three-dimensional coordinate system. and These are the rotation matrix and translation vector, respectively, projected from the 3D point cloud coordinates to the 2D image plane coordinates. For feature matching parameters, and Let be the center point of the plane of the image. After transformation, we can obtain equation (2).

[0041] (2) The heading angle calculation unit is used to calculate the heading angle of the guide vehicle. The heading angle is calculated based on the maximum and minimum distances between the current vehicle and the guide vehicle, as well as the lateral distance between the maximum and minimum points. The formula (3) describing the heading angle of the guide vehicle is shown below: (3) In the formula, and These are the distances between the currently monitored vehicle and the guide vehicle. L i The maximum and minimum values, This represents the horizontal distance between the maximum and minimum points.

[0042] when When the value is less than a certain threshold, the lead vehicle continues to travel in a straight line. Point located at Figure 7 O of the coordinates 2 X 2 Y 2 When the heading angle is positive in the first and fourth quadrants, the vehicle veers to the right. When the point is located in the second or third quadrant, the heading angle is negative, and the vehicle veers to the left.

[0043] As an optional implementation, the lateral lane-changing obstacle avoidance module uses a high-definition camera 2 to detect whether there are other traffic participants in front of the current vehicle that occupy the movement trajectory of the guide vehicle for a long time and affect the current vehicle's movement. These traffic participants are traffic obstacles, and the traffic obstacle information is sent to the central control module.

[0044] As an optional implementation, when the trajectory of the guide vehicle contains traffic obstacles, the system determines whether there are any other traffic participants within 100 meters ahead of the vehicle's left lane. If there are no other traffic participants, the central control module controls the vehicle's actuators to change lanes to the left and perform an overtaking maneuver. After completing the overtaking maneuver, the lateral lane change and obstacle avoidance module will continuously detect traffic participants in the right-front lane and return to the guide vehicle's trajectory to continue following the guide vehicle in autonomous driving. The process of lateral lane change, overtaking, and returning to the guide vehicle's original trajectory is as follows: Figure 8 As shown.

[0045] As an optional implementation, the vehicle-to-vehicle (V2V) communication module uses UWB equipment for wireless communication between vehicles. In platooning, vehicles are not completely independent; they exchange information via UWB. During platooning, the V2V communication module sends the driving status data of the guide vehicle (including throttle opening, steering wheel angle, brake pedal opening angle, and driving trajectory) to the central control module of the unmanned following vehicle, participating in the decision-making process of the unmanned following vehicle.

[0046] The vehicle-to-vehicle communication module feeds back the driving status data of the unmanned following vehicle in the convoy (including surround view information, distance, heading angle, throttle opening and closing angle, and steering wheel angle collected during the driving process) to the lead vehicle in the convoy, thereby dynamically adjusting the driving dynamics of the entire convoy in real time, and realizing stable, safe and reliable linkage control of the convoy to achieve unmanned driving.

[0047] As an optional implementation, the central control module is AGX Xavier, which receives warning information from the surround view hazard warning module, distance and heading angle from the longitudinal dynamic distance adjustment module, traffic obstacle information from the lateral lane change obstacle avoidance module, and driving status data of the guide vehicle transmitted from the vehicle-to-vehicle communication module. Based on the above information, it performs decision analysis, generates decision information, and transmits the decision information to the vehicle actuator of the unmanned following vehicle.

[0048] Meanwhile, the central control module receives driving data from the vehicle actuators, performs decision analysis again, and dynamically adjusts the driving status of the unmanned following vehicle to achieve stable and safe driving.

[0049] The working principle of the forward following linkage driving system for platoon formation provided in this embodiment is as follows: The convoy consists of a lead vehicle and multiple unmanned following vehicles. The lead vehicle sends its driving status data to the unmanned following vehicles through a vehicle-to-vehicle communication module. The driving status data includes: speed, acceleration, driving trajectory, throttle opening angle, brake opening angle, and steering wheel angle.

[0050] The central control module of the unmanned following vehicle receives driving status data sent by the guide vehicle through the vehicle-to-vehicle communication module, performs decision analysis, generates decision information, and transmits it to its own vehicle actuators to drive itself to stably follow the guide vehicle in unmanned driving.

[0051] During the operation of the unmanned following vehicle, various sensors in the vehicle actuators collect data, including steering wheel angle, accelerator pedal opening and closing degree, and pedal opening and closing degree. The collected information is fed back to the central control module for decision analysis, forming a closed-loop control to promote the stable, safe and reliable driving of the following vehicle.

[0052] During the unmanned following vehicle's operation, the longitudinal dynamic distance adjustment module monitors the distance between the current vehicle and the guide vehicle in real time, calculates the heading angle of the guide vehicle during its journey, and sends the distance and heading angle to the central control module; the lateral lane-changing obstacle avoidance module detects existing traffic obstacles on the guide vehicle's trajectory in real time and sends the traffic obstacle information to the central control module; the surround-view hazard warning module acquires environmental information around the current vehicle, determines the distance between the current vehicle and the surrounding traffic participants, and decides whether to send a warning message to the central control module based on this distance.

[0053] The central control module dynamically adjusts the driving distance between vehicles based on distance and heading angle, which is executed by the vehicle actuators; it makes decisions on lane changing, overtaking, and other actions of the current vehicle based on traffic obstacle information, which are also executed by the vehicle actuators; and it dynamically adjusts the driving status of the unmanned following vehicle based on early warning information.

[0054] The central control module sends the current driving status data of the unmanned following vehicle to the guide vehicle through the vehicle-to-vehicle communication module, realizing the coordinated control of the vehicle formation and ensuring the reliability, safety and stability of the unmanned following vehicle.

[0055] According to a second aspect of the present invention, this embodiment provides a method for implementing a forward-following linked driving system as described above for convoy formation, comprising the following steps: S1: Real-time monitoring of the environmental information around the current unmanned following vehicle and the distance between the current unmanned following vehicle and the guide vehicle; real-time detection of existing traffic obstacles on the guide vehicle's driving trajectory; establishment of communication between the current unmanned following vehicle and the guide vehicle; and exchange of information on the driving status data between the guide vehicle and the unmanned following vehicle. S2: Determine the distance between the current unmanned following vehicle and the traffic participants around it based on the environmental information monitored by S1, and decide whether to send a warning message based on the distance; Based on the maximum and minimum distances between the current unmanned following vehicle and the guide vehicle monitored by S1, as well as the lateral distances between the maximum and minimum points, the heading angle of the guide vehicle during the driving process is calculated. S3: Based on the driving status data of the guide vehicle in S1, the distance between the current unmanned following vehicle and the guide vehicle, and traffic obstacle information, as well as the warning information in S2 and the heading angle of the guide vehicle during driving, adjust the driving status of the current unmanned following vehicle, and generate driving status data of the unmanned following vehicle to feed back to the guide vehicle, so as to realize the linkage control between the guide vehicle and the unmanned following vehicle.

[0056] The driving status data of the lead vehicle includes throttle opening, steering wheel angle, brake pedal opening angle, and driving trajectory. The driving status data of the unmanned following vehicle includes environmental information, distance, heading angle, throttle opening angle, and steering wheel angle collected during the driving process.

[0057] Through the above technical solutions, the surround-view hazard warning module can monitor the environmental information around the convoy in real time and dynamically adjust the movement status of the unmanned following vehicles, thereby improving the safety performance of the unmanned convoy. The longitudinal dynamic distance adjustment module and the lateral lane-changing obstacle avoidance module acquire the movement trajectory, heading angle, vehicle-to-vehicle spacing, and traffic participants in front of the unmanned following vehicles in the convoy in real time, and simultaneously adjust the movement status of the unmanned following vehicles, improving the reliability and stability of the unmanned following vehicles' forward movement. The V2V vehicle-to-vehicle communication system promotes wireless communication of information in the convoy, enabling the convoy to adjust the movement status of the unmanned following vehicles in the convoy in real time according to traffic conditions, promoting the stability, safety, and integrity of the unmanned forward movement of the convoy. The central control module can collect real-time data transmitted from various sensors in the convoy and dynamically adjust the movement of the unmanned following vehicles through a closed-loop structure. After the integration of various systems and equipment, a complete convoy forward following system can be formed to ensure the stability, safety, and reliability of the unmanned convoy.

[0058] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A forward-following linkage driving system for convoy formation, comprising a lead vehicle and multiple unmanned following vehicles following the lead vehicle, characterized in that, The unmanned following vehicle is equipped with: Surround view hazard warning module: used to obtain environmental information around the current vehicle, determine the distance between the current vehicle and the traffic participants around the current vehicle based on the environmental information, and determine whether to send a warning information to the central control module based on the distance; Longitudinal dynamic distance adjustment module: used to measure the distance between the current vehicle and the guide vehicle in real time, calculate the heading angle of the guide vehicle during the driving process, and send the distance and heading angle information to the central control module; Lateral lane-changing obstacle avoidance module: used to detect existing traffic obstacles on the guide vehicle's trajectory in real time and send the traffic obstacle detection information to the central control module; Vehicle-to-vehicle communication module: used to establish communication between the unmanned following vehicle and the guide vehicle, and to exchange information on the driving status data between the guide vehicle and the unmanned following vehicle; Central control module: Used to receive and analyze warning information, the current distance between the vehicle and the guide vehicle, the heading angle of the guide vehicle, traffic obstacle information and the driving status data of the guide vehicle, and send the decision information to the vehicle actuators; Vehicle actuators: used to receive and execute decision information from the central control module; The surround view hazard warning module includes: Surround view monitoring unit: used to monitor the environmental information around the current vehicle; Target detection unit: Used to detect targets based on environmental information, and to detect traffic participants around the current vehicle; Hazard warning unit: Used to determine whether the distance between the current vehicle and surrounding traffic participants is less than the safe distance; if so, it sends a warning message to the central control module. The longitudinal dynamic distance adjustment module includes: Distance measurement unit: Used to monitor the distance between the current vehicle and all traffic participants within a set range ahead. L ai The system uses an object detection network to detect the guide vehicle among traffic participants, extracts the guide vehicle's position information in front of the current vehicle, and then extracts the distance between the current vehicle and the guide vehicle. L i ; Heading angle calculation unit: used to calculate the heading angle of the guide vehicle. The heading angle is calculated based on the maximum and minimum distances between the current vehicle and the guide vehicle, as well as the lateral distance between the maximum and minimum points. The distance measurement unit obtains the distance between the lead vehicle and the unmanned following vehicle in the fleet formation by data fusion of the hybrid solid-state radar (3) and the high-definition camera (2). The hybrid solid-state radar (3) and the high-definition camera (2) are arranged side by side on the bracket (4). The bracket (4) is located at the center of the head of the current vehicle. The hybrid solid-state radar (3) is used to collect the three-dimensional spatial point cloud information of the vehicle. The high-definition camera (2) is used to obtain the pixel coordinates of the lead vehicle in the two-dimensional image. The lateral lane-changing obstacle avoidance module detects in real time whether there are traffic obstacles on the driving trajectory of the guide vehicle. When there are traffic obstacles, it determines whether there are traffic participants within the set range of the left lane of the current vehicle. When there are no traffic participants, it sends information to the central control module. The central control module controls the vehicle actuator to drive the vehicle to change lanes to the left lane and perform overtaking actions. After completing the overtaking maneuver, the lateral lane-changing obstacle avoidance module detects in real time whether there are any traffic participants in the right front lane of the current vehicle. If there are no traffic participants, the central control module controls the vehicle actuator to drive the vehicle back to the movement trajectory of the guide vehicle and continue to follow the guide vehicle for unmanned driving.

2. The forward-following linkage driving system for platoon formation according to claim 1, characterized in that, The surround view monitoring unit includes multiple fisheye cameras (1) installed on the current vehicle. The spherical images acquired by the multiple fisheye cameras (1) are transformed into a square field of view through projection transformation, thereby obtaining the environmental information of the current vehicle.

3. The forward-following linkage driving system for platoon formation according to claim 2, characterized in that, The target detection unit uses Yolov3 as the backbone network, ResNet16 as the encoder, and standard classification cross-entropy and binary entropy loss functions as loss functions to construct and train the target detection model. The environmental information is input into the target detection model to detect traffic participants around the current vehicle.

4. The forward-following linkage driving system for platoon formation according to claim 1, characterized in that, The formula describing the heading angle of the guide vehicle is as follows: In the formula, and These are the distances between the currently monitored vehicle and the guide vehicle. L i The maximum and minimum values, This represents the horizontal distance between the maximum and minimum points.

5. A method for implementing a forward-following linked driving system for platoon formation as described in any one of claims 1 to 4, characterized in that, Includes the following steps: S1: Real-time monitoring of the environmental information around the current unmanned following vehicle and the distance between the current unmanned following vehicle and the guide vehicle; real-time detection of existing traffic obstacles on the guide vehicle's driving trajectory; establishment of communication between the current unmanned following vehicle and the guide vehicle; and exchange of information on the driving status data between the guide vehicle and the unmanned following vehicle. S2: Determine the distance between the current unmanned following vehicle and the traffic participants around it based on the environmental information monitored by S1, and decide whether to send a warning message based on the distance; Based on the maximum and minimum distances between the current unmanned following vehicle and the guide vehicle monitored by S1, as well as the lateral distances between the maximum and minimum points, the heading angle of the guide vehicle during the driving process is calculated. S3: Based on the driving status data of the guide vehicle in S1, the distance between the current unmanned following vehicle and the guide vehicle, and traffic obstacle information, as well as the warning information in S2 and the heading angle of the guide vehicle during driving, adjust the driving status of the current unmanned following vehicle, and generate driving status data of the unmanned following vehicle to feed back to the guide vehicle, so as to realize the linkage control between the guide vehicle and the unmanned following vehicle.

6. The linkage driving method according to claim 5, characterized in that, The driving status data of the guide vehicle includes throttle opening, steering wheel angle, brake pedal opening angle, and driving trajectory.

Images