Lateral Maneuvering with a UAV Mitigating Lateral CG Variations: Modeling and an Efficient Adaptive Backstepping Control

Abstract

In this paper, an adaptive backstepping-based control scheme is proposed to perform autonomous lateral maneuvers under significant lateral offset in the center of gravity (c.g.) position in a UAV. It is first shown that the coupled equations of motion arising from lateral c.g. shift can be simplified and cast in block strict feedback form making it amenable to a two-step backstepping control design. Useful nonlinear terms in the equations of motion are identified and retained in the backstepping design to ensure a less conservative control. Adaptation law is incorporated to dynamically adjust to changes in the c.g. position by adding an adaptive term to each step of the backstepping control. Lyapunov’s direct method and LaSalle’s invariance principle are applied to establish asymptotic stability of both tracking errors and errors in the c.g. estimate. To validate the effectiveness of the proposed control strategy, simulation results for horizontal turn maneuver are presented for the fixed wing Aerosonde UAV and maneuver performance is observed to remain highly insensitive to a wide range of lateral c.g. positions on either side of the fuselage centerline. Furthermore, a comparative control performance analysis is carried out against an ad-hoc model-based adaptive backstepping control scheme available in the literature and the results show significant performance enhancement in the proposed scheme. Along with the c.g. variations, the effects of steady crosswind are also investigated and the control formulation is modified to mitigate these effects too. Real-time control hardware in loop simulations are also provided in support of the real time viability of the proposed control.