Time-optimal solar sail transfers from Earth to pole-sitters at Mars and Venus

Abstract

Recent studies have shown the feasibility of (quasi-)pole-sitter orbits at Mars and Venus, which involves a satellite positioned along or near the polar axis of a planet in order to have a continuous, hemispherical view of the planet's polar regions. In order to further demonstrate the feasibility of this mission concept, this thesis investigates time-optimal solar sail transfers to these (quasi-)pole-sitters. In particular, (quasi-)pole-sitters which are achievable when assuming solar sail technology expected in a near- to mid-term time-frame. To reduce mission operational cost, the objective of this research is to minimize the time required for the transfer, which requires the solution to an optimal control problem. Initial guess solutions for this optimal control problem are provided through two completely different techniques, in order to compare and validate the individual performances: first, a technique derived from dynamical systems theory (a type of grid search) and second, a genetic algorithm. Subsequent optimization using a direct pseudospectral algorithm results in time-optimal transfers to the considered Mars (quasi-)pole-sitters that span 2.61 and 2.72 years, and 1.07 and 1.19 years to the considered Venus (quasi-)pole-sitters. Effects due to variations in performance of the ideal sail, non-ideal sail properties, and Earth departure orbit are investigated. In addition, this paper demonstrates that a genetic algorithm is well suited to generate initial guesses for similar interplanetary transfers in the inner solar system. It provides initial guesses that outperform the more conventional grid search technique, in terms of feasibility of the initial guess transfers, as well as in computation time and ease of implementation.