Guidelines for Modeling and Simulation of Hybrid Systems

Keywords

Abstract

Hybrid systems are defined as systems that exhibit mixed properties of continuous-time systems, discrete-time systems, and discrete-event systems. The paper analyzes various aspects that need to be taken into consideration when modeling such systems for the purpose of a numerical simulation. It addresses issues related to the convenience of the modeler when dealing with such systems, as well as issues related to the numerical properties of the generated simulation code. It shows that the difficulties encountered in modeling such systems are quite different in nature from those encountered in simulating them. When addressed in their proper contexts, these two types of problems can almost totally be decoupled from each other. A set of guidelines for the design of hybrid modeling and simulation software is derived that may support future software designers in their attempt at developing products that are both more user-friendly in terms of their modeling capabilities and more robust in terms of their simulation capabilities.


Interested in reading the full paper? (7 pages, 699,670 bytes, pdf)


Homepage


Last modified: January 18, 2006 -- © François Cellier