The guidance law for the Space Shuttle Orbiter is required to track a refer
ence state trajectory satisfying a given drag-acceleration versus velocity
profile. Newly available nonlinear guidance laws will require complicated m
ulti-dimensional interpolation. In this paper, Sugeno approximators (hybrid
fuzzy-crisp inference systems) trained by example are used to conduct this
interpolation. Moreover, the concept of surface-tracking guidance is intro
duced in which the reference trajectory itself is updated when severe distu
rbances warrants it. Surface tracking requires an "on-line optimizer" which
is implemented using a constellation of a few blocks, two of which are mul
ti-input/multi-output Sugeno approximators. The different approximators and
guidance laws have been studied and simulated. (C) 1999 Elsevier Science L
td. All rights reserved.