Observability-Guided Autonomous Path Planning for Quadrotor Navigation in GNSS-Denied Environments

Localization of a quadrotor in an unknown environment without access to Global Navigation Satellite System (GNSS) data, utilizing Simultaneous Localization and Mapping (SLAM), has shown promising results. The integration of SLAM and the Extended Kalman Filter (EKF) provides precise estimations of both the quadrotor’s position and the terrestrial landmark locations. This paper reviews the Observability-Based Path Planning (OBPP) and evaluates its performance in comparison to Monte Carlo Path Planning (MCPP). Furthermore, due to the importance of the initialization process in path planning for pre-planned methods, the performance of OBPP is evaluated for various initial positions. Simulation results demonstrate that the proposed method offers superior accuracy and greater robustness for different initial conditions, validating its effectiveness in diverse scenarios. The robustness and adaptability of this approach make it a valuable contribution to the field of autonomous navigation, especially in environments where GNSS is unavailable. This research advances the precision and reliability of quadrotor localization and path planning.

Keywords:

Real-Time Autonomous Path Planning , Observability Degree , Mont Carlo Path Planning , Gram Matrix