Structural analysis for modeling and design of multi-energy domain dynamic systems

Abstract

Summary form only given. Modelling and design of dynamic systems play a major role in determining the closed-loop system performance. However, many design methodologies still rely on trial-and-error procedures and numerical simulations. Due to the lack of physical insights, these approaches not only defer the conceiving of better system configurations but also possibly lead to unnecessary loss of efficiency. On the other hand, the study of dynamic interactions provides a valuable guidance for achieving proper system behaviors. Thus, the relation between the interactions of physical elements and the system dynamic features is a key issue in developing efficient modelling and design procedures. In this research, an energy-based method-the bond graph is used to provide a unified representation for multienergy domain systems. By the use of its unique energy interactions and causality implications, it is possible to determine the inherent system properties at early design stages before detailed element characteristics and equations are determined. The obtained information in turn suggests feasible directions for design improvement toward better system performance. This paper addresses two important system properties: the zero dynamics and the eigenvalues from the system configuration point of view. From the results of this research, it is shown that the analysis and design of dynamic systems can be conducted in a systematic way by studying the system configurations. The proposed procedures can be easily coded and become part of a computer-aided design package.