To address challenges in UAV state estimation within complex environments, this paper introduces the H ∞ -MCUKF algorithm. This method enhances the Unscented Kalman Filter by integrating the Maximum Correntropy Criterion's interference robustness with the H ∞ paradigm's worst-case error suppression. Validated using real UAV flight data, the algorithm demonstrates superior and stable performance in estimating position, velocity, and acceleration, effectively mitigating noise oscillations found in traditional methods. By using a section of real UAV 3D flight measurement data (sampling time T = 2.0s) for processing and analysis, H ∞ -MCUKF(at the optimal γ = 10) shows the H ∞ -MCUKF algorithm reduced the mean measurement residual norm by 11.4% compared to the MCUKF. Furthermore, in simulation comparisons, it achieved the lowest root mean square error (1.2318 m) for position estimation. This study confirms the algorithm significantly improves the accuracy and reliability of UAV state estimation, presenting a promising new technique for high-precision navigation and control systems.
Zhao et al. (Tue,) studied this question.