An automatic parking system and method for a tractor based on point cloud matching positioning

Abstract

The application provides a tractor automatic parking system and method based on point cloud matching positioning, which comprises a parking space detection and point cloud matching positioning laser radar, a main hanging angle measurement laser radar, a multimedia screen and an automatic parking controller. The automatic parking controller plans an optimal path of a vehicle from a starting point to a target point according to the relative position relationship between the target parking space and the vehicle, and sends an instruction to an execution mechanism of the vehicle, so that the vehicle follows the planned optimal path and automatically parks into the target parking space under the premise that the driver monitors the environmental information in real time. A 360-degree around-view system displays the environmental information around the vehicle during parking. A collision warning and braking system opens an automatic alarm function when an ultrasonic radar detects a collision risk during parking. The application effectively helps the driver to park the tractor into a parking lot in a parking lot and reduces collision accidents caused by parking.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to an automatic parking system and method for tractors based on point cloud matching and localization. Background Technology

[0002] Truck-trailer combinations are the main type of vehicle used in modern road transport. Due to their long size, large blind spots, and the articulated connection between the tractor and trailer, truck-trailer combinations are difficult to control during parking. In a limited space, it is difficult for the driver to successfully park the truck in one go.

[0003] Therefore, existing technical solutions involve installing integrated navigation equipment and using satellite signals for vehicle positioning. The parking controller then plans a parking route based on the vehicle's current location and the coordinates of available parking spaces to control the vehicle's parking. However, these existing solutions require the pre-collection and creation of offline high-precision maps, necessitating the input of the available parking space coordinates into the parking controller beforehand. In areas with unstable satellite signals, the vehicle cannot be located and therefore cannot park. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing an automatic parking system and method for tractors based on point cloud matching and positioning.

[0005] This invention is achieved using the following technical solution:

[0006] An automatic parking system for tractors based on point cloud matching and localization includes:

[0007] The parking space detection and point cloud matching LiDAR identifies available parking spaces, establishes a coordinate system, provides the coordinates of available parking spaces, and calculates the vehicle's coordinates in real time through point cloud matching.

[0008] The main vehicle and trailer angle measurement lidar detects the two-dimensional projection of the point cloud on the front side of the trailer and fits it into a straight line. It calculates the angle between the straight line and the forward direction of the main vehicle, obtains the angle information between the main vehicle and the trailer, and outputs it to the automatic parking controller.

[0009] The multimedia screen displays information on available parking spaces, and the driver selects the target parking space on the multimedia screen.

[0010] The automatic parking controller plans the optimal path from the starting point to the target point based on the relative position of the target parking space and the vehicle. At the same time, it sends instructions to the vehicle's actuators to control the brake pedal, steering wheel and gear shift, so that the vehicle follows the planned optimal path and automatically parks in the target parking space with the driver monitoring the environmental information in real time.

[0011] The 360° surround view system allows the driver to select a target parking space, and the multimedia screen switches to the 360° surround view system interface to display information about the vehicle's surrounding environment during the parking process.

[0012] The collision warning and braking system activates an automatic alarm when the ultrasonic radar detects a risk of collision during parking. When the distance to an obstacle is less than 0.5 meters, it sends a request to decelerate and perform emergency braking via the CAN bus.

[0013] The preferred method for calculating vehicle coordinates in real time using point cloud matching is as follows:

[0014] Feature extraction: Extracting features from each point cloud;

[0015] Feature matching: Matching features extracted from different point clouds to obtain the correspondence between the features of each point cloud;

[0016] Initial transformation estimation: Based on the feature matching results, estimate the initial transformation between point clouds and roughly align them;

[0017] Point-to-point or point-to-area registration: Using the iterative nearest point algorithm, the registration accuracy is optimized based on the initial transformation between point clouds;

[0018] Vehicle coordinate calculation: Real-time calculation of the coordinate changes of available parking spaces to obtain the vehicle's coordinates in a given coordinate system.

[0019] Preferably, the left and right fenders of the cab are respectively equipped with parking space detection and point cloud matching positioning laser radars, the rear end of the cab is equipped with a main trailer angle measurement laser radar, and the rear end of the trailer is equipped with an ultrasonic radar.

[0020] The multimedia screen has a built-in automatic parking app; the automatic parking controller is located under the sleeper berth in the driver's cab; the lidar is connected to the automatic parking controller via Ethernet, the multimedia screen and ultrasonic radar are connected to the automatic parking controller via CAN bus, and the automatic parking controller is connected to the vehicle's actuators via CAN bus.

[0021] A point cloud matching and localization method for automatic parking of a tractor-trailer, applied to any of the point cloud matching and localization-based automatic parking systems for tractor-trailers described above, includes the following steps:

[0022] a. The driver drives the vehicle into the parking lot, clicks to start the automatic parking APP on the multimedia screen, and manually drives the vehicle at a constant speed of no more than 10km / h.

[0023] b. During vehicle operation, the parking space detection and point cloud matching positioning lidar projects the detected obstacle point cloud into a two-dimensional grid map. Within the target area, adjacent empty grids are connected to form a parking space to be measured. The size of the parking space to be measured is calculated to see if it is larger than a standard truck parking space. If it is larger than a standard truck parking space, it is determined to be a parking space that can be parked, and the coordinate information of the parking space is sent to the automatic parking controller and multimedia screen.

[0024] c. After the multimedia screen receives the coordinate information of the available parking space, it will display a white parking space. The driver will continue to drive the vehicle manually. When the white parking space meets the parking conditions, the white parking space on the multimedia screen will turn into a green parking space, and the driver will stop the vehicle to select the target parking space.

[0025] d. The driver selects the target parking space on the multimedia screen, the automatic parking function is activated, and the automatic parking controller plans a parking curve based on the position of the vehicle and the position of the target parking space;

[0026] e. During the parking process, the main vehicle and trailer angle measurement lidar will project the two-dimensional point cloud of the front side of the trailer into a straight line in real time, calculate the angle between the straight line and the direction of travel of the main vehicle, obtain the angle information between the main vehicle and the trailer, and output it to the automatic parking controller.

[0027] f. The automatic parking controller derives the front wheel steering angle based on the angle information between the tractor and trailer and the vehicle dynamics model analysis. The automatic parking controller converts the front wheel steering angle into a steering wheel angle command and outputs the steering wheel angle command to the steering controller to control the vehicle's steering. At the same time, the automatic parking controller uses PID to perform longitudinal control of the vehicle to achieve constant speed driving at low speeds.

[0028] g. During the automatic parking process, the multimedia screen switches to the 360-degree surround view system to display the vehicle's surrounding environment for the driver to view; when the ultrasonic radar detects a risk of collision, the automatic alarm function is activated; when an obstacle is detected to be less than 0.5 meters away, a request for deceleration is sent via the CAN bus to initiate emergency braking.

[0029] Compared with the prior art, the present invention has the following beneficial technical effects:

[0030] It can acquire available parking spaces in real time and select a suitable target parking space through the multimedia screen. It uses PID to control the vehicle longitudinally to achieve constant speed driving at low speeds. During automatic parking, the surrounding environment can be displayed through the multimedia screen. It can effectively help drivers park trailers in parking lots, reduce collisions caused by parking, and enable trailers to perform path tracking and reversing operations efficiently and safely in complex reversing environments such as areas with unstable satellite signals. Attached Figure Description

[0031] Figure 1 This is an architecture diagram of the control system of the present invention;

[0032] Figure 2 This is a schematic diagram of the parking space detection principle in the control method of the present invention;

[0033] Figure 3 This is a schematic diagram of the control method of the present invention;

[0034] Figure 4 This is a schematic diagram of the pure tracking algorithm of this invention;

[0035] Figure 5 This is a geometric diagram showing the relationship between the main vehicle and the trailer in this invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0037] Example 1

[0038] like Figure 1 As shown, an automatic parking system for tractors based on point cloud matching and localization includes:

[0039] The parking space detection and point cloud matching LiDAR identifies available parking spaces, establishes a coordinate system, provides the coordinates of available parking spaces, and calculates the vehicle's coordinates in real time through point cloud matching.

[0040] The main vehicle and trailer angle measurement lidar detects the two-dimensional projection of the point cloud on the front side of the trailer and fits it into a straight line. It calculates the angle between the straight line and the forward direction of the main vehicle, obtains the angle information between the main vehicle and the trailer, and outputs it to the automatic parking controller.

[0041] The multimedia screen displays information on available parking spaces, and the driver selects the target parking space on the multimedia screen.

[0042] The automatic parking controller plans the optimal path from the starting point to the target point based on the relative position of the target parking space and the vehicle. At the same time, it sends instructions to the vehicle's actuators to control the brake pedal, steering wheel and gear shift, so that the vehicle follows the planned optimal path and automatically parks in the target parking space with the driver monitoring the environmental information in real time.

[0043] The 360° surround view system allows the driver to select a target parking space, and the multimedia screen switches to the 360° surround view system interface to display information about the vehicle's surrounding environment during the parking process.

[0044] The collision warning and braking system activates an automatic alarm when the ultrasonic radar detects a risk of collision during parking. When the distance to an obstacle is less than 0.5 meters, it sends a request to decelerate and perform emergency braking via the CAN bus.

[0045] The method for calculating vehicle coordinates in real time using point cloud matching is as follows:

[0046] Feature extraction: Extracting features from each point cloud;

[0047] Feature matching: Matching features extracted from different point clouds to obtain the correspondence between the features of each point cloud;

[0048] Initial transformation estimation: Based on the feature matching results, estimate the initial transformation between point clouds and roughly align them;

[0049] Point-to-point or point-to-area registration: Using the iterative nearest point algorithm, the registration accuracy is optimized based on the initial transformation between point clouds;

[0050] Vehicle coordinate calculation: Real-time calculation of the coordinate changes of available parking spaces to obtain the vehicle's coordinates in a given coordinate system.

[0051] The left and right fenders of the cab are respectively equipped with laser radar for vehicle position detection and point cloud matching positioning, and the rear of the cab is equipped with laser radar for measuring the main trailer angle. The rear of the trailer is equipped with ultrasonic radar.

[0052] The multimedia screen has a built-in automatic parking app; the automatic parking controller is located under the sleeper berth in the driver's cab; the lidar is connected to the automatic parking controller via Ethernet, the multimedia screen and ultrasonic radar are connected to the automatic parking controller via CAN bus, and the automatic parking controller is connected to the vehicle's actuators via CAN bus.

[0053] Example 2

[0054] like Figures 2-3 The method for automatic parking of a tractor-trailer based on point cloud matching and localization, as shown, is applied to any of the point cloud matching and localization-based automatic parking systems for tractor-trailers described above, and includes the following steps:

[0055] a. The driver drives the vehicle into the parking lot, clicks to start the automatic parking APP on the multimedia screen, and manually drives the vehicle at a constant speed of no more than 10km / h.

[0056] b. During vehicle operation, the parking space detection and point cloud matching positioning lidar projects the detected obstacle point cloud into a two-dimensional grid map. Within the target area, adjacent empty grids are connected to form a parking space to be measured. The size of the parking space to be measured is calculated to see if it is larger than a standard truck parking space. If it is larger than a standard truck parking space, it is determined to be a parking space that can be parked, and the coordinate information of the parking space is sent to the automatic parking controller and multimedia screen.

[0057] c. After the multimedia screen receives the coordinate information of the available parking space, it will display a white parking space. The driver will continue to drive the vehicle manually. When the white parking space meets the parking conditions, the white parking space on the multimedia screen will turn into a green parking space, and the driver will stop the vehicle to select the target parking space.

[0058] d. The driver selects the target parking space on the multimedia screen, the automatic parking function is activated, and the automatic parking controller plans a parking curve based on the position of the vehicle and the position of the target parking space;

[0059] e. During the parking process, the main vehicle and trailer angle measurement lidar will project the two-dimensional point cloud of the front side of the trailer into a straight line in real time, calculate the angle between the straight line and the direction of travel of the main vehicle, obtain the angle information between the main vehicle and the trailer, and output it to the automatic parking controller.

[0060] like Figure 4 As shown in the diagram, in the pure tracking algorithm... The aiming distance is determined by the intersection of the aiming radius arc and the target trajectory, which is the aiming point. This is determined by the change in the trailer's angle. The target master-slave angle can be derived. .

[0061] The geometric relationship between the tractor and trailer is as follows: Figure 5 As shown, the following relationship is derived between the front wheel steering angle and the main suspension angle:

[0062] ;

[0063] in, The turning radius of the main vehicle. It is the turning radius of the trailer;

[0064] ;

[0065] Further derivation yields the following method for calculating the expected main hanging angle:

[0066] ;

[0067] The front wheel angle of the target is calculated based on the difference between the expected main wheel angle and the actual measured main wheel angle, as shown below, where k is the feedback gain coefficient.

[0068] .

[0069] f. The automatic parking controller derives the front wheel steering angle based on the angle information between the tractor and trailer and the vehicle dynamics model analysis. The automatic parking controller converts the front wheel steering angle into a steering wheel angle command and outputs the steering wheel angle command to the steering controller to control the vehicle's steering. At the same time, the automatic parking controller uses PID to perform longitudinal control of the vehicle to achieve constant speed driving at low speeds.

[0070] g. During the automatic parking process, the multimedia screen switches to the 360-degree surround view system to display the vehicle's surrounding environment for the driver to view; when the ultrasonic radar detects a risk of collision, the automatic alarm function is activated; when an obstacle is detected to be less than 0.5 meters away, a request for deceleration is sent via the CAN bus to initiate emergency braking.

Claims

1. An automatic parking system for tractors based on point cloud matching and localization, characterized in that, include: The parking space detection and point cloud matching LiDAR identifies available parking spaces, establishes a coordinate system, provides the coordinates of available parking spaces, and calculates the vehicle's coordinates in real time through point cloud matching. The main vehicle and trailer angle measurement lidar detects the two-dimensional projection of the point cloud on the front side of the trailer and fits it into a straight line. It calculates the angle between the straight line and the forward direction of the main vehicle, obtains the angle information between the main vehicle and the trailer, and outputs it to the automatic parking controller. The multimedia screen displays information on available parking spaces, and the driver selects the target parking space on the multimedia screen. The automatic parking controller plans the optimal path from the starting point to the target point based on the relative position of the target parking space and the vehicle. At the same time, it sends instructions to the vehicle's actuators to control the brake pedal, steering wheel and gear shift, so that the vehicle follows the planned optimal path and automatically parks in the target parking space with the driver monitoring the environmental information in real time. The 360° surround view system allows the driver to select a target parking space, and the multimedia screen switches to the 360° surround view system interface to display information about the vehicle's surrounding environment during the parking process. The collision warning and braking system activates an automatic alarm when the ultrasonic radar detects a risk of collision during parking. When the distance to an obstacle is less than 0.5 meters, it sends a request to decelerate and perform emergency braking via the CAN bus.

2. The automatic parking system for tractor vehicles based on point cloud matching and positioning according to claim 1, characterized in that, The method for calculating the vehicle's coordinates in real time using point cloud matching is as follows: Feature extraction: Extracting features from each point cloud; Feature matching: Matching features extracted from different point clouds to obtain the correspondence between the features of each point cloud; Initial transformation estimation: Based on the feature matching results, estimate the initial transformation between point clouds and roughly align them; Point-to-point or point-to-area registration: Using the iterative nearest point algorithm, the registration accuracy is optimized based on the initial transformation between point clouds; Vehicle coordinate calculation: Real-time calculation of the coordinate changes of available parking spaces to obtain the vehicle's coordinates in a given coordinate system.

3. The automatic parking system for tractor vehicles based on point cloud matching and positioning according to claim 1, characterized in that, The left and right fenders of the cab are respectively equipped with LiDAR for vehicle position detection and point cloud matching positioning, and the rear of the cab is equipped with LiDAR for measuring the main trailer angle. The rear of the trailer is equipped with ultrasonic radar. The multimedia screen has a built-in automatic parking app; the automatic parking controller is located under the sleeper berth in the driver's cab; the lidar is connected to the automatic parking controller via Ethernet, the multimedia screen and ultrasonic radar are connected to the automatic parking controller via CAN bus, and the automatic parking controller is connected to the vehicle's actuators via CAN bus.

4. A point cloud matching and positioning method for automatic parking of a tractor, applied to the point cloud matching and positioning-based automatic parking system for tractors as described in any one of claims 1-3, characterized in that, Includes the following steps: a. The driver drives the vehicle into the parking lot, clicks to start the automatic parking APP on the multimedia screen, and manually drives the vehicle at a constant speed of no more than 10km / h. b. During vehicle operation, the parking space detection and point cloud matching positioning lidar projects the detected obstacle point cloud into a two-dimensional grid map. Within the target area, adjacent empty grids are connected to form a parking space to be measured. The size of the parking space to be measured is calculated to see if it is larger than a standard truck parking space. If it is larger than a standard truck parking space, it is determined to be a parking space that can be parked, and the coordinate information of the parking space is sent to the automatic parking controller and multimedia screen. c. After the multimedia screen receives the coordinate information of the available parking space, it will display a white parking space. The driver will continue to drive the vehicle manually. When the white parking space meets the parking conditions, the white parking space on the multimedia screen will turn into a green parking space, and the driver will stop the vehicle to select the target parking space. d. The driver selects the target parking space on the multimedia screen, and the automatic parking function is activated. The automatic parking controller plans a parking curve based on the vehicle's position and the target parking space position. e. During the parking process, the main vehicle and trailer angle measurement lidar will project the two-dimensional point cloud of the front side of the trailer into a straight line in real time, calculate the angle between the straight line and the direction of travel of the main vehicle, obtain the angle information between the main vehicle and the trailer, and output it to the automatic parking controller. f. The automatic parking controller derives the front wheel steering angle based on the angle information between the tractor and trailer and the vehicle dynamics model analysis. The automatic parking controller converts the front wheel steering angle into a steering wheel angle command and outputs the steering wheel angle command to the steering controller to control the vehicle's steering. At the same time, the automatic parking controller uses PID to perform longitudinal control of the vehicle to achieve constant speed driving at low speeds. g. During the automatic parking process, the multimedia screen switches to the 360-degree surround view system to display the vehicle's surrounding environment for the driver to view; when the ultrasonic radar detects a risk of collision, the automatic alarm function is activated; when an obstacle is detected to be less than 0.5 meters away, a request for deceleration is sent via the CAN bus to initiate emergency braking.

Images