Efficient astrodynamics-informed kinodynamic motion planning for relative spacecraft motion

In this paper, we present a computationally efficient sampling-based spacecraft relative motion planning algorithm to reconfigure a spacecraft from a given initial state to a desired final state while avoiding collisions with obstacles and minimizing the total fuel required for the transfer. In our previous work, we presented an astrodynamics-informed kinodynamic motion planning (AIKMP) algorithm that can regularly find fuel-efficient and collision-free transfer solutions for relative spacecraft motion – without requiring an initial guess of the solution. In this work, we improve the computation efficiency of the AIKMP algorithm. We demonstrate that by using a linearized Lambert solution (LLS) instead of a full Lambert solution as the steering law for the developed motion planning algorithm, along with an additional tree-pruning module, the computation efficiency of the AIKMP algorithm can be improved by ≈96%. The new pruning module added to the AIKMP algorithm also significantly reduces the storage requirement by the algorithm. Using different example relative transfer scenarios comprising two spacecraft and multiple obstacles/keep-out zones, we demonstrate the performance of the proposed efficient kinodynamic orbital motion planner and show that the algorithm efficiently computes near-optimal, fuel-efficient, and collision-free trajectories for spacecraft relative transfer problems.