|
Simulating Quadrotor Dynamics in the Presence of Constant Winds |
|
Minhyeok Kwon, Yongsoon Eun(DGIST, Korea) |
| This paper analyzes dynamics of quadrotor in a wind field using a recently developed model. In particular, equiibria under wind have been analyzed, and trajectory tracking performance under LQR controller in the presence of wind is presented. |
|
|
Learning-Based NMPC for Agile Navigation and Obstacle Avoidance in Unstructured Environments |
|
Oualid Doukhi, Daeuk kang, Yoonha Ryu, Jaeho Lee, Deok Jin Lee(Jeonbuk National University, Korea) |
| Autonomous drones operating in unstructured environments face challenges that demand agile navigation and obstacle avoidance. However, the lack of robust vision-based algorithms has hindered the development of effective control strategies. This paper presents a novel approach that combines Learning-Based NMPC (Nonlinear Model Predictive Control) with deep learning policies to address this problem. In our approach, a deep learning policy is trained using an expert NMPC that has access to the environment's map in simulation. The trained policy is then integrated with the NMPC as a feed-forward control signal, ensuring obstacle avoidance while reaching the desired waypoints. |
|
|
Automatic Transition Maneuvers of Tilt-Rotor Aircraft within Conversion Corridor |
|
Heetae Park, Hoijo Jeong, Daeun Hwang, Junhyun Lim, Jinyoung Suk, Seungkeun Kim(Chungnam National University, Korea) |
| This paper covers the automatic transition maneuvers of a tilt-rotor aircraft within the conversion corridor. The trim envelope is calculated for a given flight condition in level flight, and the reference transition line and conversion corridor are determined to minimize the pitch angle deviation during the transition maneuvers. The nonlinear dynamic inversion-based controllers are designed to regulate pitch angle and velocity in the direction of the tilt angle. Lastly, we conducted numerical simulation for forward and backward transition and demonstrated that the tilt-rotor aircraft could safely achieve transition within the conversion corridor. |
|
|
GoLive - A modular Mixed Reality Simulation for Semantic Plug and Play |
|
Constantin Wanninger, Thomas Badem, Martin Schörner, Christian Eymüller, Alexander Poeppel, Wolfgang Reif(University of Augsburg, Germany) |
| GoLive is not only a new type of modular simulation
but in combination with a MR visualization it adds value
to the development of robotics applications and the move
from simulation to reality. MR devices serve both as
a tangible user interface for rapid restructuring of scenarios and as the first physical test setup of a decentralized deployment. The focus here is primarily on mobile robot applications developed with the ROS ecosystem.
By decoupling the visualization and the simulation models, not only purely functional simulations with a practical graphical interface can be constructed, but also illustrative demonstrators with special requirements for different end devices. |
|
|
Indoor Pedestrian-Following System by a Drone with Edge Computing and Neural Networks: Part 1 - System Design |
|
In-Chan Ryu, Jung-Il Ham, Jun-Oh Park, Jae-Woo Joeng(Gwangju Institute of Science and Technology, Korea), Sung-Chang Kim(Electronics and Telecommunications Research Institute, Korea), Hyo-Sung Ahn(GIST, Korea) |
| As the drone market continues to expand, the need for accurately determining a drone’s position and orientation
using a camera in GPS-denied environments becomes increasingly critical. This paper aims to achieve precise position
and attitude data by incorporating SLAM to provide visual measurements for EKF, thereby ensuring the stability of drone
operations. An experiment was conducted to execute commands from the ground control PC using the map created as a
result of SLAM. The primary tools used for this purpose included the Pixhawk Orange, Jetson Nano, and the ZED-Mini
camera. The research showcases the effectiveness of these tools and methods in enhancing indoor drone functionality. |
|
|
Indoor Pedestrian-Following System by a Drone with Edge Computing and Neural Networks: Part 2 - Development of Tracking System and Monocular Depth Estimation |
|
Jungil Ham, InChan Ryu, Jun Oh Park, JaeWoo Jeong(GIST, Korea), Sung-Chang Kim(Electronics and Telecommunications Research Institute, Korea), Hyo-Sung Ahn(GIST, Korea) |
| This paper is the second installment in a series on indoor drone pedestrian tracking utilizing edge computing
and neural networks. Building upon the SLAM and EKF technologies introduced in Part 1, this paper introduces Monocular Depth Estimation to reduce camera costs and overall weight. The system leverages AI-driven depth information for
indoor positioning and real-time human tracking. Experiments demonstrate the drone’s ability to autonomously track a
specific individual indoors using vision and IMU sensors. Key contributions encompass an AI-based tracking system
employing YOLO v3 and a novel depth estimation approach that supersedes traditional depth cameras. |
|