Simulation and Flight Test of Autonomous Aircraft Estimation, Planning, and Control Algorithms

Abstract

The development, summary, and validation of real time autonomous estimation, planning, and control algorithms for nonlinear aircraft are presented. Two square root estimation algorithms are developed and used to estimate the state dynamics, parameters, and uncertainty bounds of a nonlinear aircraft in real time. A square root sigma point filter approximates the uncertain state using a finite set of points, while a square root extended set-membership filter bounds the states and parameters of the nonlinear system using ellipsoidal sets. The estimation methods are integrated into online optimized model predictive control methodology for fast reconfigurable control, and a path planner based on streamlines for fast path planning around obstacles and for target pursuit. The performances of the estimation, planning, and control techniques are tested in flight using the SeaScan autonomous aircraft. The SeaScan platform is briefly described, and real time hardware in the loop and flight data are presented for a variety of applications using each of the algorithms.