Qualitative Simulation of Nonlinear Dynamical Systems
Antonio Vicino · 2005
The aim of this paper is to investigate potentialities of qualitative simulation teeniques to study the behavior of nonlinear dynamical systems from system theory point of view. Among the different qualitative reasoning theories recently proposed in the literature, we focus on the approach of Kuipers. The basic QSIM algorithm proposed by Kuipers appears to have desiderable features for the qualitative analysis of dynamical sys- tems we are interested in. In particular, a first investigation on nonlinear autonomous systems is carried out and to this purpose the QSIM algorithm has been implemented showing how additional rules for global filtering can be used for reducing possible spurious or apparently different behaviors. Particular attention has been devoted to the recognition of typical patterns found in dynamical systems behavior, such as cycles and equilibrium points. The simulation algorithm has been implemented on a Vax 8800 computer in a Poplog environment. The results of qualitative simulation of some nonlinear dynamical systems are presented.