Koopman-Based Dynamic Environment Prediction for Safe UAV Navigation

TL;DR

This paper introduces a Koopman-based model predictive control framework for UAVs that predicts moving obstacles in real-time, enabling safe navigation through dynamic environments with robustness to noise and delays.

Contribution

It develops a novel Koopman operator approach integrated with MPC for real-time dynamic obstacle prediction in UAV navigation, validated through simulation and ROS2-Gazebo.

Findings

Reliable obstacle prediction under sensor noise

Robust trajectory planning despite actuation delays

Effective real-time navigation in dynamic environments

Abstract

This paper presents a Koopman-based model predictive control (MPC) framework for safe UAV navigation in dynamic environments using real-time LiDAR data. By leveraging the Koopman operator to linearly approximate the dynamics of surrounding objets, we enable efficient and accurate prediction of the position of moving obstacles. Embedding this into an MPC formulation ensures robust, collision-free trajectory planning suitable for real-time execution. The method is validated through simulation and ROS2-Gazebo implementation, demonstrating reliable performance under sensor noise, actuation delays, and environmental uncertainty.