Autonomous Constrained Control of Arbitrarily-Configured Spacecraft with Broad Uncertainties

Current spacecraft control methods rely on explicit, predetermined commands, making them computationally intensive and inflexible for vehicles with mass properties and thruster configurations that change over time due to changing cargo configurations or thruster malfunctions. This work introduces a methodology within special Euclidean group SE(3) to autonomously control any spacecraft with constrained thruster configurations to enable autonomous spacecraft control under uncertainty. The approach integrates an unscented Kalman filter (UKF) on tangent bundle of SE(3), i.e. TSE(3), and a Morse-Lyapunov controller to determine optimal thruster firing sequences while overcoming state, measurement, and process uncertainties. Simulations demonstrate successful stationkeeping under randomized thruster layouts and mass distributions, successfully enabling stationkeeping for any configuration of mass properties and thrusters. Autonomous docking is successfully performed in simulations under sensor and dynamic uncertainties, including state estimation errors, measurement noise and alignment errors, and thruster misalignment. Results validate the robustness of the method in mitigating state, actuator, and sensor uncertainties, with applications to autonomous GN&C of constrained spacecraft configurations considering saturation limits.