Year: 2025
Type: conference-paper
Abstract: Unmanned Aerial Vehicles (UAVs) require robust perception and control strategies for autonomous tracking of moving targets. This work integrates Extended Kalman Filter (EKF)-based trajectory prediction with a flatness-based Model Predictive Control (MPC) scheme for vision-based tracking. The EKF smooths noisy measurements and predicts short-term target motion, ensuring continuity during occlusions. These predictions serve as references for an MPC controller formulated on the quadrotor’s differentially flat outputs, enabling a simplified linear predictive model while respecting acceleration and input constraints. Real-time simulations demonstrate accurate and smooth tracking of the EKF-predicted trajectories, confirming the effectiveness of the combined EKF-MPC framework for vision-guided UAV target tracking.
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7127438158