ROS2 robot system-based unmanned aerial vehicle tractor path planning method and system

Abstract

The invention discloses an unmanned aerial vehicle tractor path planning method and system based on an ROS2 robot system, the system comprises an RK3566 development board, a laser radar, a shell and a depth vision camera, the RK3566 development board is provided with an ROS2 system, and the laser radar, the development board and the depth camera are installed in the shell. According to the present invention, the use effect is good, the ROS2 system installed on the RK3566 development board is adopted to perform rtabmap vision and laser radar outdoor 3D mapping navigation so as to achieve the autonomous navigation and the automatic driving of the tractor, and the vision camera is adopted to detect and identify the mark and the target so as to substantially improve the automation degree, the maneuverability and the operation effect of the aircraft tractor.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of aircraft towing vehicles, in particular to an unmanned aircraft towing vehicle path planning method and system based on a ROS2 robot system. BACKGROUND

[0002] The unmanned aircraft towing vehicle path planning method and system based on the ROS2 robot system can effectively realize the cooperative work between the unmanned aircraft and the towing vehicle, and provide an efficient and flexible solution for various tasks. The system has wide application potential and will play an important role in the fields of automated logistics, agriculture, environmental monitoring and the like. The system realizes the communication and coordination between the unmanned aircraft and the towing vehicle by utilizing the ROS2 framework, and generates an efficient path by using a path planning algorithm. The architecture of the system mainly includes the following components: mapping and environment perception, before the unmanned aircraft and the towing vehicle start a task, the environment is modeled and perceived by using a perception device (such as a laser radar and a camera); path generation, based on the environment map and the task requirements, a proper path planning algorithm (such as an A algorithm and a D algorithm) is used to generate the path of the unmanned aircraft towing vehicle; path optimization, the generated path is optimized to make the path smoother and more efficient, so as to improve the overall performance of the system; path tracking, in the actual execution process, the unmanned aircraft and the towing vehicle need to be controlled according to the generated path, and the heading, speed and other parameters are adjusted in real time, so as to realize accurate path tracking. SUMMARY

[0003] In order to realize the automatic update and control of the aircraft towing vehicle, the application provides an unmanned aircraft towing vehicle path planning method and system based on a ROS2 robot system. The system allows the aircraft towing vehicle to reach the correct predetermined position under the rtabmap visual system and the laser radar 3D mapping after receiving the operation command, and the characteristics include: RTAB-Map SLAM mapping, a ROS2 system, a laser radar, a depth camera and an RK3566 development board.

[0004] The RTAB-Map SLAM mapping is based on the ROS2 operating system, installed on a Linux board, using an RK3566 chip as the main control chip. RTAB-Map (Real-Time Appearance-Based Mapping) is an open-source augmented reality and visual SLAM (Simultaneous Localization and Mapping) system used to build and maintain environment maps. SLAM refers to the task of simultaneously localizing a robot's own position and creating a map by fusing perception data. RTAB-Map can utilize sensor information such as RGB-D cameras or laser radars, as well as robot motion information, to perform real-time environment modeling and self-localization. It uses an appearance-based approach to extract objects and scene features in the environment as visual words, and updates the map and localization based on the combination of these words. This algorithm can handle not only dense environments but also large-scale environments, and has been optimized for loop detection and loop optimization.

[0005] The ROS2 system ROS2 is "Robot Operating System 2", an open-source platform for building robot systems. It is the next generation of ROS (Robot Operating System) and aims to provide a more powerful, reliable and flexible framework for robot software development.

[0006] The laser radar is a device that uses laser technology for measurement and perception. It emits pulsed laser beams and receives reflected laser signals to obtain information about the position, shape, speed and other related information of target objects. Laser radar has many advantages, such as high precision, long measurement range, high resolution, fast response, etc. It has a wide range of applications in the fields of autonomous driving, robot navigation, three-dimensional modeling, environmental perception, etc. For example, laser radar can be used to create a three-dimensional map of the vehicle's surroundings and provide accurate obstacle detection and distance measurement, helping autonomous vehicles drive safely. In addition, in the industrial field, laser radar can also be used to measure and monitor the position and shape of target objects.

[0007] The depth camera is a special camera that can perceive the distance and depth information of objects in a scene. It uses different techniques than traditional cameras, such as structured light, time-of-flight, or stereo vision, to obtain more accurate depth information. Depth cameras are commonly used in computer vision applications, such as virtual reality, augmented reality, pose recognition, and 3D reconstruction. By obtaining depth information in a scene, depth cameras can provide a more realistic and immersive interactive experience. In virtual reality, depth cameras can track the user's movements and position and reflect them in the virtual environment, enabling interaction with virtual objects. The working principle of a depth camera varies depending on the specific technology used. For example, a structured light camera uses a special infrared laser or light pattern to illuminate the scene, then captures the reflected image through a camera, and uses the distortion of the light to calculate the distance of objects. Another common technique is the time-of-flight camera, which sends out a pulse of light and calculates the time it takes for the light to reflect back from the camera to the object, thus obtaining distance information.

[0008] The RK3566 development board is an embedded system development board released by Rockchip. It is based on the RK3566 processor, which uses ARM architecture and has high performance and low power consumption characteristics. RK3566 integrates four Cortex-A55 CPU cores and one Cortex-A76 CPU core, providing excellent multi-threaded processing capabilities. In addition, it also integrates a Mali-G52 2EE GPU, supporting high-quality graphics processing. RK3566 supports 8K H.264 / H.265 hardware decoding and encoding, as well as 4K VP9 hardware decoding. This allows it to handle high-resolution video streams and perform well in multimedia applications. DETAILED DESCRIPTION

[0009] The rodless tractor starts and the system automatically initializes and self-checks, and the indicator light lights up;

[0010] The depth camera and laser radar data are transmitted to the RK3566 development board for SLAM mapping, and calculation is performed through the established 3D model;

[0011] The input tractor information before running is used to calculate the tractor rotation radius and front wheel rotation angle of the tractor.

[0012] The present application has the following advantages:

[0013] An unmanned aerial vehicle tractor path planning method and system based on a ROS2 robot system are adopted to perform automatic operation of an airport aircraft tractor, so that the airport operation is more stable and the automation degree is higher. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 is the working process of the aircraft tractor path planning method and system of the present application; DETAILED DESCRIPTION

[0015] The technical solutions of the present application are further described below, but the scope of protection is not limited to the description.

[0016] Example one:

[0017] As shown in the drawings, Figure 1 The present application is an unmanned aerial vehicle tractor path planning method and system based on a ROS2 robot system, which is loaded on an aircraft rodless tractor. After receiving the operation command, the system allows the aircraft rodless tractor to reach the correct predetermined position under the rtabmap vision system and laser radar 3D mapping, including: RTAB-Map SLAM mapping, ROS2 system, laser radar, depth camera, RK3566 development board; The system can plan the route during driving and can avoid obstacles by itself.

[0018] The working principle is that the depth camera and the laser radar serve the RK3566 development board. When the development board receives a positioning request, RTAB-Map SLAM mapping is performed, the rotation radius of the tractor and the front wheel rotation angle of the tractor are calculated, and the aircraft tractor is guided to automatically map, navigate and avoid obstacles.

Claims

1. ROS2-based robot system unmanned aerial vehicle tractor path planning method and system, the path planning method and system are applied to an aircraft rodless tractor, the system realizes autonomous navigation and automatic obstacle avoidance of the tractor. The system allows the aircraft rodless tractor to reach the correct predetermined position under the rtabmap vision system and laser radar 3D mapping after receiving the operation command, characterized in that it comprises: RTAB-MapSLAM mapping, ROS2 system, lidar, depth camera, RK3566 development board. 2.The unmanned aerial vehicle towing vehicle path planning method and system of claim 1, wherein: The rtabmap vision system, RTAB-Map (Real-Time Appearance-Based Mapping) is an open-source software library for visual SLAM (Simultaneous Localization and Mapping). It can construct a three-dimensional map of the environment in real-time by using camera and sensor data, while simultaneously localizing the robot. RTAB-Map can handle changes in the scene in dynamic environments, and has functions such as loop closure detection, keyframe selection, and scene reconstruction. This software library is suitable for many application fields, such as robot navigation, augmented reality, virtual reality, etc. 3.The unmanned aerial vehicle towing vehicle path planning method and system of claim 1, wherein: The ROS2 system, ROS2 is the second generation of the Robot Operating System. Compared with ROS1, ROS2 introduces some new features and improvements to provide a more powerful and flexible robot development framework. 4.The unmanned aerial vehicle tow truck path planning method and system of claim 1, wherein: The lidar is a technology that detects targets by emitting and receiving laser signals. It uses a laser beam to scan the target and determines its position, shape, and motion state by analyzing the reflection or scattering of the laser. Lidar is widely used in autonomous vehicles, unmanned aerial vehicle navigation, weather observation, etc.

5. The unmanned aerial vehicle tow vehicle path planning method and system of claim 1, wherein: The RK3566 development board uses RK3566 main control chip, carries Linux operating system, can install ROS2 system, has GPU or FPU type calculation acceleration module for AI deep learning acceleration; has CAN bus and serial communication protocol implementation and laser radar depth camera and the communication requirements of the lower machine. RK3566 chip is a high integration, low power system chip launched by Ruiyi Micro. The chip uses advanced 28nm process, has four Cortex-A55 cores and a NPU (neural network processor), supports up to 8GB LPDDR4x memory and 1080p display resolution. RK3566 chip has strong computing and image processing capabilities, suitable for smart home, industrial automation, smart TV and other fields. It supports Android and Linux operating systems, and provides rich interfaces and functions such as multimedia decoding, audio processing, face recognition, etc. 6.The method and system for path planning of UAV tow vehicle according to claim 1, wherein: Depth camera, also known as three-dimensional camera or RGB-D camera, is a camera that can acquire the distance and depth information of objects in a scene. Traditional cameras can only obtain two-dimensional color images, while depth cameras can capture the distance value of each pixel point to the camera. This makes depth cameras play an important role in many application fields, such as robot navigation, virtual reality, augmented reality, three-dimensional modeling, etc. Depth cameras usually use structured light, time-of-flight, binocular vision and other technologies to measure distance. Among them, structured light cameras emit special light sources and observe the deformation of light patterns on the surface of objects to calculate depth information; time-of-flight cameras measure the time difference between the emission of pulses from the camera and the reception of reflected light by the receiver to calculate depth; binocular vision cameras acquire images at different angles through two lenses of the camera, and calculate the depth by calculating the parallax between the two angles. Depth cameras have wide applications, such as in the field of robots, they can be used for environmental perception, obstacle detection and obstacle avoidance, etc.; in the field of virtual reality and augmented reality, depth cameras can capture users' actions in real time and apply them to virtual scenes, thus providing a more immersive experience; at the same time, depth cameras are also used for three-dimensional modeling and scanning to quickly obtain the geometric shape of target objects.

Images