II. DIAGNOSIS SYSTEM USING ARTIFICIAL
INTELLIGENCE TECHNIQUES
Most diagnosis systems are implemented using expert
systems, in which basic models based on analytical models
are combined with those based on quality based models. This
rationale is the basis for the elaboration of the architecture of
such a system – Fig. 1. - which combines heuristic knowledge
obtained through experiments, with qualitative and
quantitative knowledge regarding the model [1][2].
Command and diagnosis units represent the central part of
the system, integrating methods based on analytical models
which use empirical information (qualitative model) as well
as methods based on intimate knowledge of the system [3][4].