Robust Sampling-Based Covariance Steering for Aerocapture Guidance

摘要

Aerocapture is a maneuver where a spacecraft dives through the atmosphere of a planet or moon to reduce its velocity and prepare for orbital insertion. Aerocapture allows for higher cruise velocities and reduces fuel consumption, decreasing transit time and increasing payload mass. However, uncertainties in the atmospheric entry state and atmospheric density increase the risk of aerocapture. Dynamic nonlinearities and nonlinearities caused by the state-dependence of the atmospheric density pose additional challenges. This work develops a robust sampling-based covariance steering algorithm designed for aerocapture guidance. Our proposed algorithm leverages sampled nonlinear system trajectories to improve evaluation of the delta-V required for aerocapture and address nonlinearities caused by the aerocapture dynamics and atmospheric disturbances. We perform Monte Carlo simulations with dispersed entry and atmospheric conditions on aerocapture scenarios at Mars and Uranus and demonstrate a 5-15% reduction in the 99th-percentile, 99.7th-percentile, and worst-case delta-V required for aerocapture when compared against a state-of-the-art covariance steering algorithm.