Comparative analysis of Kalman filtering methods in tightly-coupled GNSS-INS integration for UAV navigation systems

This paper presents a comprehensive comparative study of Kalman filtering algorithms (EKF, UKF, CKF) in tightly-coupled GNSS-INS integrated navigation for UAVs. Based on Simulation and experimental results encompassing six characteristic flight regimes, we systematically evaluate their performance in dynamic adaptability, computational efficiency, and environmental robustness. The measured results agree well with theoretical predictions, which exhibit superior performance such as conventional scenarios (hovering, low-speed tracking), significant performance divergence emerges under high-dynamic maneuvers (emergency avoidance, rapid ascent) and GNSS-degraded conditions (urban canyons, signal outages) In this paper, the characteristics of three kinds of filters are analyzed based on engineering experience, which represents a significant contribution for the development and application of UAV integrated navigation technology in the future.