Neural network-based sliding mode variable structure control for Mars entry

Abstract

To deliver a Mars entry vehicle through the Martian atmosphere to the prescribed parachute deployment point, active Mars entry guidance and control is essential. This article addresses the problem of Mars atmospheric entry control by a neural network-based sliding mode variable structure control (NNSMVSC) to reduce the effect of the bounded uncertainties on the atmospheric density and aerodynamic coefficients. First, NNSMVSC is designed to robustly track the prescribed nominal trajectory under high uncertainties and to effectively reduce the downrange error. Then, the heading alignment logic is adopted in the lateral plane to reduce the cross-range error. Finally, the validity of the control algorithm proposed in this article is demonstrated by computer simulation analysis.