Constrained adaptive backstepping take-off control for a morphing hybrid aerial underwater vehicle

Abstract

This paper studies the inclined take-off control for a kind of morphing hybrid aerial underwater vehicles. At the initial stage, the hybrid aerial underwater vehicle could not achieve trim condition for the low speed. Then the altitude could have an obvious decrease during the take-off process. This could lead the vehicle to crashing into water. To escape from the unbalanced stage, an adaptive backstepping controller is designed to track the target angle of attack and the thrust is opened to maximum within limitation. The radial basis function Neural Networks and parameter adaption are employed in order to deal with the time-varying nonlinear variable sweep dynamics of the vehicle. Bisides, the control input of the elevator deflection is easy to be saturated for the low initial velocity. This is taken into account in the design of a constraint adaptive backstepping controller to prevent aggressive adaptions and to provide reliable input commands. Simulations are presented to varify the proposed controller.