Behavioral interpretation according to multimodeling representation

This paper deals with a behavioral interpretation of physical systems. It d escribes how we can interpret the systems current state using a multimodeli ng principle. The purpose of this interpretation task is to explain what is happening in the system from a small set of observations in terms of funct ional states of the system processes. The model types considered are functi onal and qualitative models generated according to a bond graph description . The functional model (process model) is contructed taking into account th e causality and the orientation of the energy flow considered in the bond g raph. We cooperate this model with a qualitative causal graph in order to i dentify the system process functional states.