Robust Trajectory Optimization for Guided Powered Descent and Landing

A robust trajectory optimization approach for guidance algorithm gain selection
for powered descent and landing is developed. This approach uses a genetic algorithm
to determine optimal guidance algorithm parameters while incorporating
uncertainty information from linear covariance analysis. The optimal guidance algorithm
parameters are determined while accounting for environment, navigation,
and vehicle property uncertainty and sensor suite fidelity. As a demonstration of
this method, the optimal gains for the fractional polynomial powered descent guidance
are found for the braking phase of a robotic lunar landing mission. Scenarios
with differing sensor suites and sensor qualities are considered, with objective
functions to minimize variability in propellant usage or terminal position. Results
show that the optimal guidance algorithm gains for a given trajectory differ based
on the sensor suite, and optimal guidance algorithm gains may result in up to 20%
performance improvements over the baseline in propellant usage and landed accuracy.