Gain indexing schemes for low-thrust perturbation guidance

Abstract

Two perturbation guidance schemes, time-to-go guidance and minimum-distance guidance, are re-examined in the context of a low-thrust orbit transfer problem. The two schemes, which use different techniques for indexing feedback gains, are shown to be comparable in performance. Both schemes are found to produce terminal state errors that are orders of magnitude smaller than those obtained in several previous studies. Various small modifications or enhancements of the algorithms are thought to account for a portion of the dramatic improvement in results. The problem investigated is a hypothetical Earth to Mars orbit transfer with six state variables, two control variables, and six terminal state constraints.