Finite-Thrust Natural-Motion Circumnavigation Injection by Convex Optimization
Ping Lu, Alexander Lewis, Richard J. Adams, Michael D. DeVore, Christopher Petersen
Abstract
A methodology is developed to use convex optimization for finding the propellant-optimal finite-thrust trajectory of a spacecraft to inject into a specified natural-motion circumnavigation (NMC) orbit around another spacecraft. The problem is nonconvex. A philosophically new perspective is introduced to take advantage of modern convex optimization. Through a novel analysis the NMC problem is shown to be equivalent to a two-dimensional constrained optimization problem. This conceptually simpler interpretation enables two numerical approaches to be investigated, one based on convex relaxation and the other linearization-projection. It is established that while each approach is able to lead to the solution to the NMC-injection problem in a subset of the possible cases, their domains of applicability complement each other to cover all possible cases. A hybrid algorithm is designed that combines the strengths of the two approaches and enables the application of convex optimization to solve the NMC-injection problem in all cases where the solution exists, without the need for any user-supplied parameters or initial guesses. The effectiveness of the hybrid algorithm is demonstrated in finding the numerical solutions to the NMC problem reliably and rapidly.