Airplane rodless tractor automatic wheel holding system based on FAST-LIO2
By adopting an automatic wheel holding system based on FAST-LIO2 in rodless tractors, combined with radar inertia odometer and filtering algorithm, the automated docking and wheel holding tasks of aircraft rodless tractors are realized, and it is difficult for drivers to accurately adjust the body. Position issues improve traction efficiency and safety.
Patent Information
- Application Number
- CN202310983501.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-07
- Publication Date
- 2025-02-18
AI Technical Summary
When the rodless tractor is picked up, it is difficult for the driver to adjust the position of the vehicle body accurately, resulting in uneven force on the left and right rollers when the wheel holder clamps the front wheel of the aircraft, which can easily cause sudden deflection during the traction process, damage the aircraft landing gear, and Drivers who are not skilled in driving can lead to a decrease in traction efficiency.
Adopt the automatic wheel-holding system of the aircraft rodless tractor based on FAST-LIO2, combining the radar inertia odometer frame, filter, upper computer module, lower computer module and sensor group to realize automated connection with the aircraft wheel-holding and ground environment map construction. . The distance information of the aircraft's front wheel or main landing gear is obtained through lidar, and combined with filtering algorithms and sensor data to achieve efficient and accurate aircraft ground operation.
The automated docking and wheel holding tasks of aircraft rodless tractors have been realized, traction efficiency and safety have been improved, the complexity and human error of the pilot's operation have been reduced, and the safety of the aircraft landing gear has been ensured.
Smart Images
Figure CN119460140A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of aircraft tractors, and in particular to an automatic wheel-holding system of an aircraft rodless tractor based on FAST-LIO2. Background Art
[0002] An aircraft tractor is a support equipment used to tow aircraft on the ground at an airport. When an aircraft is being manufactured, repaired, or its engine is shut down, the movement of the aircraft requires special towing equipment. With the development of towing technology, two types of tractors have been developed: towbar tractors and towbarless tractors. Towbarless tractors directly clamp the front wheels of an aircraft and lift the front landing gear of the aircraft. They are special vehicles used to tow aircraft. Towbarless tractors have the following problems: when the tractor driver docks the aircraft, he does not adjust the position of the vehicle body to completely align with the front wheels of the aircraft. In this case, when the wheel-holding device clamps the front wheels of the aircraft, the forces on the left and right rollers are uneven, which can easily lead to sudden deflection during towing and damage the aircraft landing gear. Drivers with less skilled driving skills are slow in adjusting the center line of the vehicle body to align with the front wheels of the aircraft, resulting in reduced towing efficiency. Summary of the invention
[0003] In order to solve the above technical problems, the present invention proposes an automatic wheel holding system of an aircraft rodless tractor based on FAST-LIO2, including a FAST-LIO2 radar inertial odometer framework, a filter, a host computer module, a slave computer module, and a sensor group.
[0004] The SLAM mapping based on FAST-LIO2 realizes automatic wheel connection with the aircraft and builds a real-time ground environment map. When the system is running, the laser radar is responsible for obtaining the distance information of the aircraft's front wheels or main landing gear, while scanning the surrounding environmental point cloud data. Through real-time processing of the laser radar data, fast laser odometer and positioning of ground vehicles are realized. Efficient and accurate aircraft ground operations are achieved.
[0005] The filtering algorithm can accurately estimate the map of the environment and the position of the robot. Filtering is a technique used to remove noise, smooth data, and estimate unknown variables. The following filtering schemes are used in combination, including Kalman filtering and particle filtering: Particle filtering is a non-parametric filtering technique based on the Monte Carlo method.
[0006] The host computer uses AMD or ARM main control chip, uses Linux operating system, can be equipped with ROS system, has GPU or FPU type computing acceleration module for AI deep learning acceleration; has CAN bus and serial port 485 communication protocol to realize the communication requirements with laser radar depth camera and lower computer.
[0008] The lower computer can realize the servo control of the wheel motor, can communicate with the posture sensor to collect and solve data, can receive the command of the upper computer and transmit the acquired data to the upper computer. The servo wheel motor is driven by pure electricity. The motor is installed on the side of the wheel to drive the wheel alone. The motor is directly driven by the motor reducer, which can realize fine control of the wheel torque and torque. It can realize zero-radius in-situ turning, forward, backward, and precise turning through the four-wheel differential. It can accurately execute the instructions sent by the lower computer.
[0009] The sensor module mainly includes LiDAR and Inertial Measurement Unit (IMU). FAST-LIO2 achieves fast robot positioning and mapping by effectively fusing LiDAR and IMU data, which is especially suitable for mobile linear vector chassis systems with limited computing resources. The data provided by LiDAR is dense and accurate, while IMU provides fast-response motion information. The fusion of the two can overcome the limitations of each when used alone, making FAST-LIO2 an efficient and accurate positioning and mapping solution. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Figure 1 This is a schematic diagram of the wheel-holding system of the aircraft rodless tractor of the present invention; DETAILED DESCRIPTION
[0011] The technical solution of the present invention is further described below, but the scope of protection claimed is not limited to the description.
[0012] Embodiment 1:
[0013] In order to solve the above technical problems, the present invention proposes an automatic wheel holding system for a rodless aircraft tractor, which automatically searches for the aircraft landing gear wheelset and automatically completes the docking and wheel holding tasks. During this period, the position and speed parameters of the tractor, the real-time position and posture of the tractor and the aircraft landing gear, and the surrounding environment information can be displayed on the screen in real time. The system can complete the docking and wheel holding tasks by itself, greatly improving efficiency and safety.
[0014] The technical solution is as follows:
[0015] The present invention discloses a FAST-LIO2 radar inertial odometer framework, a filter, a host computer module, a slave computer module and a sensor group.
[0016] The FAST-LIO2 technology combines the automatic wheel holding of the aircraft barless tractor and the SLAM mapping system to realize an efficient and accurate aircraft ground operation solution. In this integrated system, laser radar and odometer become key sensors to help ground vehicles automatically connect with the aircraft to achieve wheel holding and build a real-time ground environment map. When the system is running, the laser radar is responsible for obtaining the distance information of the aircraft's front wheels or main landing gear, while scanning the surrounding environmental point cloud data. FAST-LIO2 technology plays an important role. Through the real-time processing of laser radar data, the fast laser odometer and positioning of ground vehicles are realized. The efficiency and accuracy of FAST-LIO2 ensure that ground vehicles can accurately track the changes in the position and attitude of the aircraft. The automatic control system calculates the appropriate contact point and wheel holding position based on the position information provided by FAST-LIO2 and the preset wheel holding algorithm, thereby realizing the automatic wheel holding connection between the ground vehicle and the aircraft. At the same time, FAST-LIO2 technology can not only be used to realize the automatic wheel holding function, but also to build a point cloud map of the ground environment in real time. Ground vehicles perform SLAM mapping based on LiDAR data, continuously updating maps and reflecting changes in aircraft paths and airport ground environments. These maps not only help improve the positioning accuracy of ground vehicles, but also provide ground operators with valuable information for optimizing ground paths and improving aircraft ground operation efficiency.
[0017] In radar mapping, in order to accurately estimate the map of the environment and the position of the vehicle, it is usually necessary to filter the sensor data. Filtering is a technique used to remove noise, smooth data, and estimate unknown variables. In radar mapping, common sensor filtering methods include: Kalman filter: Kalman filter is the most common filter and is widely used in radar mapping. It is a recursive state estimation filter that estimates the optimal state of the system based on the predicted state and sensor measurements. Kalman filter works very well in the case of Gaussian noise and can effectively estimate the position and map of the vehicle. Extended Kalman filter (EKF): EKF is an extended version of Kalman filter, which is suitable for nonlinear systems and non-Gaussian noise. In radar mapping, EKF is often used to estimate the state and map of the robot due to the presence of nonlinear characteristics and non-Gaussian noise of the sensor. Unscented Kalman filter (UKF): UKF is an improvement on EKF, which estimates the state of nonlinear systems more accurately by using unscented transformation. UKF avoids the error caused by linearization in EKF, so it can better handle nonlinear problems in radar mapping. Particle Filter: Particle filter is a non-parametric filtering technology based on the Monte Carlo method. It represents the vehicle state and map by randomly sampling a set of particles, and then updates the weights based on the sensor data. Particle filtering performs well in nonlinear and non-Gaussian noise environments and can better cope with multimodal problems. These filtering methods can be selected and adjusted according to specific application scenarios and sensor characteristics to improve the accuracy and stability of radar mapping. In practical applications, the appropriate filtering method is usually selected based on the characteristics of the sensor and the noise model to meet the requirements of mapping.
[0018] The lower computer can realize the servo control of the wheel motor, can communicate with the posture sensor to collect and solve data, can receive the command of the upper computer and transmit the acquired data to the upper computer. The servo wheel motor is driven by pure electricity. The motor is installed on the side of the wheel to drive the wheel alone. The motor is directly driven by the motor reducer, which can realize fine control of the wheel torque and torque. It can realize zero-radius in-situ turning, forward, backward, and precise turning through the four-wheel differential. It can accurately execute the instructions sent by the lower computer.
[0019] The host computer uses AMD or ARM main control chip, uses Linux operating system, can be equipped with ROS system, has GPU or FPU type computing acceleration module for AI deep learning acceleration; has CAN bus and serial port 485 communication protocol to realize the communication requirements with laser radar depth camera and lower computer.
[0020] The sensors mainly include laser radar and inertial measurement unit (IMU). Laser radar: Laser radar is the core sensor of FAST-LIO2, which is used to perceive the three-dimensional point cloud data of the surrounding environment. It determines the distance and position of surrounding objects by emitting laser beams and measuring the reflected signals returned. Laser radar can provide high-precision distance information and can obtain map data in real time in indoor and outdoor environments. In FAST-LIO2, the data of laser radar is used to build maps, extract features, and fuse with IMU data to achieve high-precision odometer estimation. Inertial measurement unit (IMU): IMU is a sensor combination, usually including accelerometers and gyroscopes, which are used to measure the linear acceleration and angular velocity of the robot. IMU can provide the robot's posture information and acceleration information for estimating the robot's motion state. In FAST-LIO2, the data of IMU is used to estimate the robot's posture change and acceleration, and then combined with the data of laser radar for efficient laser odometer estimation. FAST-LIO2 achieves fast robot positioning and mapping by effectively fusing laser radar and IMU data, which is particularly suitable for mobile robot systems with limited computing resources. Because the data provided by LiDAR is dense and accurate, while the IMU provides fast-response motion information, the fusion of the two can overcome the limitations of each when used alone, making FAST-LIO2 an efficient and accurate mobile robot positioning and mapping solution.
Claims
1. An automatic wheel-holding system for a towbarless aircraft tractor based on FAST-LIO2. The automatic wheel-holding system for a towbarless aircraft tractor is applied to a towbarless aircraft tractor. The invention allows the use of the FAST-LIO2 framework to complete spatial modeling and complete the automatic wheel-holding of the towbarless aircraft tractor. Its features include a FAST-LIO2 radar inertial odometer framework, a filter, a host computer module, a slave computer module, and a sensor group.
2. The automatic wheel holding system for a towbarless aircraft tractor according to claim 1 is characterized in that: The ROS-based route planning system is based on the ROS operating system, odometer, nine-axis IMU, laser radar, and depth camera. The ROS system is used to establish real-time route planning on the map and is loaded on the embedded host computer. The nine-axis IMU is used to obtain the position and posture of the tractor and the real-time changes of the road surface. The embedded lower computer reads through the SPI bus and then calculates and processes and transmits the data to the embedded host computer for storage; the laser radar is used for real-time distance measurement, using laser to scan the 360° range around the barless tractor, and transmits the data back to the host computer ROS system to complete SLAM. The depth camera is used for precise distance measurement and image recognition of the landing gear. Through feature point comparison and depth map correction, the aircraft enters the camera recognition range to accurately navigate from the barless tractor to the docking position in front of the aircraft landing gear.
3. The automatic wheel holding system for a towbarless aircraft tractor according to claim 1 is characterized in that: The AI machine vision posture solving system takes real-time images of the aircraft landing gear, collects feature point information, and transmits the information to the embedded host computer through the CAN bus. The information is processed by combining the traditional CV machine vision algorithm and the AI deep learning algorithm to achieve more accurate and rapid feature extraction, thereby solving the analytical space coordinates and deflection angle posture state information of the aircraft landing gear and tractor. The data obtained by the machine vision is also uploaded to the docker of the embedded host computer for training to enhance the AI deep learning effect.
4. The automatic wheel holding system for a towbarless aircraft tractor according to claim 1 is characterized in that: The embedded lower computer can realize servo control of the wheel-side motor, can communicate with the posture sensor to collect and solve data, can receive instructions from the upper computer and transmit acquired data to the upper computer. It has the following capabilities: eight-way CAN bus communication capability, four-way serial port communication protocol, four-way I2C communication protocol, eight-way 24V voltage output capability, eight-way 24V voltage input capability, and ST-link interface. It supports online programming of the host computer and can read and write to the host computer through the serial port. The embedded lower computer and the embedded host computer establish a connection through the CAN bus to achieve fast communication, and communicate with slower requirements through the serial port.
5. The automatic wheel holding system for a towbarless aircraft tractor according to claim 1 is characterized in that: The embedded host computer uses an AMD or ARM main control chip, uses a Linux operating system, can be equipped with a ROS system, and has a GPU or FPU-type computing acceleration module for AI deep learning acceleration; it has a CAN bus and serial port communication protocol to achieve communication requirements with the laser radar depth camera and the lower computer.
6. The rodless automatic wheel holding system for a rodless aircraft tractor according to claim 1, characterized in that: The servo wheel motor is driven by pure electricity. The motor is installed on the side of the wheel to drive the wheel alone. The motor is directly driven by the motor reducer, which can achieve fine control of wheel torque and torque. It can achieve zero-radius in-situ turning, forward, backward, and precise turning through the four-wheel differential.
7. The automatic wheel holding system for a towbarless aircraft tractor according to claim 1 is characterized in that: The wheel holding system is a functional component of the rodless tractor, and is composed of a wheel holding mechanism, a hydraulic system and a control system. When the omnidirectional mobile rodless tractor is aligned with the aircraft landing gear wheelset, the control system starts the hydraulic pump station after obtaining the clamping instruction, and controls the hydraulic valve group to open and close the drive of each hydraulic cylinder of the wheel holding mechanism to complete the clamping and releasing action of the wheel holding mechanism. After the wheel holding mechanism clamps and lifts the aircraft landing gear wheelset, the aircraft landing gear wheelset is lifted off the ground, and the omnidirectional mobile rodless tractor can tow the aircraft to the designated location.