Near-Optimal Explicit Guidance and Reduced-Attitude Control for Lunar Descent and Pinpoint Landing

Abstract This study proposes and tests a new guidance and control architecture for powered descent and pinpoint landing at the lunar South Pole. Guidance employs three subsequent explicit schemes: (1) Lambert-based finite-thrust guidance, (2) locally flat near-optimal guidance, and (3) predictive bang-bang vertical guidance. These are based on either the Lambert theorem, optimal control applied to simplified dynamics, or prediction along the final vertical path. The attitude control system has the final goal of aligning the actual thrust direction with the desired one, provided by the guidance algorithm. The resulting reduced-attitude control problem is addressed through a recently introduced quaternion-based nonlinear control algorithm, which has been proven to enjoy asymptotic stability properties. The attitude actuation system is composed of 12 monopropellant thrusters, ignited using pulse width modulation, in conjunction with an array of single-gimbal control momentum gyroscopes. The spacecraft dynamics is propagated in a high-fidelity dynamical framework, with inclusion of several relevant orbit perturbations. A Monte Carlo campaign with nonnominal flight conditions testifies to the effectiveness of the guidance and control architecture at hand, in terms of propellant consumption and precision at landing.