Multi-Objective Robust Trajectory Design for Powered Descent and Landing
Grace E. Calkins, Zachary R. Putnam, David Woffinden · 2024
A robust trajectory optimization approach is proposed for powered descent and landing by selecting optimal guidance algorithm gain and target vector. The method employs a genetic algorithm to minimize uncertainties arising from environment, navigation, and vehicle properties on flight performance, considering a specific sensor suite. Vehicle state uncertainties are efficiently computed using linear covariance analysis techniques. The approach shapes a trajectory favorable to the navigation sensor suite, leading to improved flight performance. To demonstrate the effectiveness of this method, optimal guidance parameters are determined for a multi-phase trajectory of a robotic lunar landing mission. Results indicate that the selected parameters outperform baseline ones in terms of propellant usage and terminal accuracy. Multi-objective optimization reveals the trade-off between terminal position uncertainty and total propellant usage for various sensor suite compositions. Moreover, the guidance parameters chosen through this method may enable acceptable flight performance with fewer or lower-quality sensors.