Towards Autonomous Operation of UAVs Using Data-Driven Target Tracking and Dynamic, Distributed Path Planning Methods

A hybrid real-time path planning method has been developed that employs data-driven target UAV trajectory tracking methods. It aims to autonomously manage the distributed operation of multiple UAVs in dynamically changing environments. The target tracking methods include a Gaussian mixture model, a...

Full description

Saved in:
Bibliographic Details
Published in:Aerospace 2024-09, Vol.11 (9), p.720
Main Authors: Choi, Jae-Young, Prasad, Rachit, Choi, Seongim
Format: Article
Language:English
Subjects:
Algorithms
Aviation
Changing environments
Cloud computing
Collision avoidance
Communication
Computational linguistics
data-driven target tracking
Distributed memory
distributed path planning
Environmental changes
Extended Kalman filter
Free flight
Kalman filters
Language processing
Machine learning
Methods
Natural language interfaces
Neural networks
Operations management
Path planning
Planning
Potential fields
Probabilistic models
R&D
Real time
Research & development
Rotary wing aircraft
Software
software-in-the-loop
Tracking
Trajectory planning
UAS dynamic
Unmanned aerial vehicles
UTM
Vehicles
Virtual networks
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A hybrid real-time path planning method has been developed that employs data-driven target UAV trajectory tracking methods. It aims to autonomously manage the distributed operation of multiple UAVs in dynamically changing environments. The target tracking methods include a Gaussian mixture model, a long short-term memory network, and extended Kalman filters with pre-specified motion models. Real-time vehicle-to-vehicle communication is assumed through a cloud-based system, enabling virtual, dynamic local networks to facilitate the high demand of vehicles in airspace. The method generates optimal paths by adaptively employing the dynamic A* algorithm and the artificial potential field method, with minimum snap trajectory smoothing to enhance path trackability during real flights. For validation, software-in-the-loop testing is performed in a dynamic environment composed of multiple quadrotors. The results demonstrate the framework’s ability to generate real-time, collision-free flight paths at low computational costs.
ISSN:2226-4310
2226-4310
DOI:10.3390/aerospace11090720