Risk-aware Mission Design for In situ Asteroid Exploration under Uncertainty

In-situ robotic asteroid exploration is integral to current and future space programs for sustainable human exploration and resupply in deep space, deflection of potentially hazardous objects, and study of the early solar system formulation. Meanwhile, dynamical environments around asteroids are highly perturbed and uncertain, and pose challenges for spacecraft to safely navigate around them. The dynamics are influenced by a variety of sources of uncertainty, such as asteroid properties, exogenous disturbances, and errors associated with operations. These uncertainties need to be properly quantified and taken into account of the mission design processes. On the other hand, characteristics of such uncertainties are not usually given or fixed but rather dependent on the mission architecture and operational scenarios. To aid in the decision-making processes with trade-offs between the mission feasibility and uncertainties, mission designers need to quantify the feasibilities of a number of architectures for a range of possible combinations of uncertainty characteristics. To explore the solution space efficiently and reliably, this paper presents a systematic approach that leverages techniques from the fields of stochastic optimal control and convex optimization. Formulated as a convex optimization problem, the solution method enables us to solve many number (as many as ~ 10
⁵
) of stochastic optimal control problems without initial guesses. The proposed approach is applied to the design of asteroid global-mapping campaigns, which demonstrates the effectiveness and validity of our approach. The result reveals important trade-off relationships between the mission feasibility and assumed uncertainty characteristics in the context of asteroid global mapping. While the convex formulation involves a dynamical approximation, the validity of the convex-programming-based solutions is confirmed through nonlinear Monte-Carlo simulations under the original dynamics.