Spacecraft Trajectory Design Using Data-Driven Model Predictive Control

This paper focuses on the optimal trajectory design in the circular restricted three-body problem with the aid of powerful data-driven approach, which can address the global nonlinear dynamics. First, the datadriven approach to approximate a Koopman operator of the nonlinear dynamical system is developed and show the constant of the motion can be restored from the time-series data of trajectories. Then two control method exploiting the reconstructed model are derived: the state-dependent Riccati equation approach and port-Hamiltonian control approach. Then, a new method to to approximate a Koopman operator without a priori knowledge about the basis functions is proposed and effectiveness of the algorithm is demonstrated.