Stochastic Atmosphere Modeling for Risk Adverse Aerocapture Guidance

In this paper we consider the problem of aerocapture guidance inside an atmosphere with the density randomly perturbed according to a Gaussian process model. We formulate the problem of solving for a single bank event as a stochastic optimization problem and then, in terms of a 3-DoF Martian aerocapture example, we study the properties of the proposed approach compared to a standard deterministic approach. The stochastic approach is shown to decrease the risk that a large delta- $V$ is required for maneuvering after atmospheric flight to reach a target apoapsis and periapsis.