Probabilistic Trajectory Optimization Under Uncertain Path Constraints for Close Proximity Operations

Abstract

Rendezvous and docking operations have been an integral component of manned spaceflight from the beginning of the space age. However, there is now a growing interest in close proximity spacecraft operations in the new areas of orbital debris removal, on-orbit assembly, on-orbit refueling, and on-orbit servicing and repair missions. Current trajectory planning methods focus on performing rendezvous and docking to very well-known targets and in very well-known conditions. Inherent to these new mission types, however, is an increasing element of uncertainty to which new trajectory optimization architectures will need to be robust. There is an inherent tradeoff here between safety and performance. This paper attempts to address the uncertainties underlying path constraints while maintaining a probabilistically optimal level of performance. The goal is to find baseline trajectories with the best expected performance over large uncertainties in mission critical parameters, such as knowledge of an obstacle’s position or attitude of a target spacecraft while knowing that the trajectory is able to be replanned on board the spacecraft when higher-precision information is obtained. Results specific to this problem and the associated analysis motivate the use of this probabilistic planning framework in future space missions.