SCSC2003 Abstract S01064

Automated Symbolic Generation of Vector Graph Theoretic Object Based Non-Linear Dynamic Simulation Models

Automated Symbolic Generation of Vector Graph Theoretic Object Based Non-Linear Dynamic Simulation Models

Submitting Author: Dr. Sriprakash Sarathy

Developing a comprehensive simulation model of a multi-body system (MBS) is not a trivial task, however, over the years many tools and techniques have been developed to make this task easy and less cumbersome. Much of the innovations and advances have come in three areas; (i)modeling interface, (ii)efficient numerical formulations, and (iii)enhanced computational schemes. With the advent of powerful computational capabilities, much of the emphasis has shifted away from modeling innovation to enhanced computational approaches. The ability to turn a model on its head and examine it form an analytical perspective has yielded to powerful visualization tools that make sense of reams of numerical data generated by complex codes. Yet, the essential role of a simulation model, namely to provide a manageable representation of reality that can be useful in comprehending the phenomena, has not changed one iota.

The work described in this paper, returns, in a sense, to the well trodden path of symbolic modeling and provides a methodology that can provide symbolic simulation models of real-life dynamic systems. To make the latter manageable, it is essential to incorporate some measure of automation in generation of such models, as well as object orientation in the modeling interface. The use of vector graph theoretic modeling applied to non-linear systems coupled with a object orientation is a useful and significant method to achieve the aforementioned goal. In addition to these advantages, inherent parallelism of these techniques makes it useful for real time physical simulations.

To do this, the authors have drawn upon a significant body of prior work in the area of vector-graph theoretic formulations for non-linear systems and coupled it with advances in physically based object oriented modeling paradigms. An example of a multi-body system is used to illustrate the automated procedures used and the symbolic models generated from it.

Back to SCSC2003 Abstracts