Search algorithms for space tug rendezvous: simulation and experiment

Abstract

Space tugs must be able to search for and identify inert non-communicative targets, whose ephemeris is only known approximately. The goal of this article is to evaluate the efficiency of three possible two-dimensional search strategies in the context of autonomous rendezvous in space. As part of a broader Space Tug project, a number of modeling challenges were addressed to validate the experimental results. A crucial problem was to reduce the complex space search problem to a simple two-dimensional search for an inert target using LEGO robots. The data collected, as well as results from simulations, showed strong trends in the relationship between time and energy expended during the search. The project provided a starting point for the rendezvous control system to be implemented in a Space Tug vehicle by showing which of the three strategies - random, semi-autonomous and autonomous - is most efficient in the two-dimensional search case. It was found that the semi-autonomous algorithm is the most energetically efficient approach, but also the most time-consuming. This finding disproved our initial belief that the semi-autonomous (human-in-the-loop) strategy is the most efficient in terms of both time and energy. Instead, the conclusion is that an autonomous algorithm is more suitable for space applications. The results also suggest that, depending on knowledge of the search space and mission requirements, a hybrid approach might be most efficient.