Continuity and change (activity) are fundamentally related in DEVS simulation of continuous systems

Abstract

The success of DEVS methods for simulating large continuous models calls for more in-depth examination of the applicability of discrete events in modeling continuous phenomena. We present a concept of event set and an associated measure of activity that fundamentally characterize discrete representation of continuous behavior. This metric captures the underlying intuition of continuity as well as providing a direct measure of the computational work needed to represent continuity on a digital computer. We discuss several application possibilities beyond high performance simulation such as data compression, digital filtering, and soft computation. Perhaps most fundamentally we suggest the possibility of dispensing with the mysteries of traditional calculus to revolutionize the prevailing educational paradigm. © Springer-Verlag Berlin Heidelberg 2005.