Adaptive trajectory tracking for UAV guidance with bayesian filtering

The objective of this paper is to design algorithms for tracking the trajectory of unmanned aerial vehicle (UAV). A reference path is generated and recorded by flight simulator software, X-Plane, which provides the flight information to guide the UAV towards the target path. The flight motion is modeled with linear ordinary differential equations, considering both longitudinal and lateral motion. The stability derivatives and aerodynamic coefficients of dynamic equations are derived from flying situations and specifications of UAV and airfoil. Based on the flight dynamics, Bayesian filtering is applied to estimate the control inputs of dynamic equations. The proposed algorithm is verified via simulations for takeoff, level flight and landing, which show that the proposed scheme is feasible for UAV guidance.