Minimization of GNSS-Denied Inertial Navigation Errors for Fixed Wing Autonomous Unmanned Air Vehicles

TL;DR

This paper introduces an enhanced inertial navigation algorithm for fixed wing UAVs that maintains accurate positioning during GNSS signal loss by integrating additional onboard sensors and validated through high-fidelity simulations.

Contribution

The paper presents a novel sensor fusion algorithm that reduces navigation errors in GNSS-denied conditions and is optimized for low SWaP UAVs, with open-source implementation.

Findings

The proposed filter outperforms traditional methods in GNSS-denied scenarios.

Sensor integration significantly reduces position and attitude errors.

Simulation results demonstrate robustness to sensor quality variations.

Abstract

This article proposes an inertial navigation algorithm intended to lower the negative consequences of the absence of GNSS (Global Navigation Satellite System) signals on the navigation of autonomous fixed wing low SWaP (Size, Weight, and Power) UAVs (Unmanned Air Vehicles). In addition to accelerometers and gyroscopes, the filter takes advantage of sensors usually present onboard these platforms, such as magnetometers, Pitot tube, and air vanes, and aims to minimize the attitude error and reduce the position drift (both horizontal and vertical) with the dual objective of improving the aircraft GNSS-Denied inertial navigation capabilities as well as facilitating the fusion of the inertial filter with visual odometry algorithms. Stochastic high fidelity Monte Carlo simulations of two representative scenarios involving the loss of GNSS signals are employed to evaluate the results, compare the proposed filter with more traditional implementations, and analyze the sensitivity of the results to the quality of the onboard sensors. The author releases the C++ implementation of both the navigation filter and the high fidelity simulation as open-source software.