Building Intelligence into an Autopilot
Using Qualitative Simulation to Support Global Decision Making
Abstract
This paper represents an addendum to an earlier paper published by the same
research group in Simulation
[Vesanterä and Cellier, 1989].
In the earlier paper, a crisp inductive reasoner had been employed to
qualitatively reason about the behavior of a quantitatively simulated B--747
aircraft, to determine when a structural malfunction occurs, to hypothesize
about the nature of this malfunction, and to decide upon a global strategy that
allows to operate the quantitative aircraft model under the modified flying
conditions.
In the addendum, the formerly crisp inductive reasoner has been replaced by a
fuzzy inductive reasoner. The paper demonstrates the enhanced discriminatory
power and the improved forecasting capability of the modified reasoning
scheme. In addition, the fuzzy inductive reasoner allows to predict a
quasi-continuous response spectrum, whereas the crisp inductive reasoner was
able to predict discrete (class) values only.
Interested in reading the
full paper?
(11 pages, 1,064,848 bytes, pdf)
Homepage
Last modified: January 12, 2006 -- © François Cellier