Metamodelling: Bond Graphs and Dynamic Systems
P.J. Gawthrop, L.S.P.S. Smith
Abstract
This is a 1996 textbook. With the increasing complexity of processes to be analysed, the<br> modern control engineer often needs to develop a model of the system<br> to be controlled. However, in many cases, there is limited time for<br> detailed system analysis, and the engineer may not be an expert in<br> that particular system domain. This book is aimed at graduate<br> engineers (and postgraduate students) who wish to use a systematic<br> approach to model development that is suited to computer-aided<br> modelling environments. The goal of this book is to support the use of modelling as a<br> useful knowledge-enhancing exercise, and to propose<br> corresponding modelling methodologies. The motivation for this<br> is the widespread use of models in analysing and simulating<br> systems for safe and cost-effective evaluation of new<br> processes. The context is primarily that of control system<br> design, due to the extensive use of models of the process, and<br> its disturbances, in modern design methods. We use the term metamodelling to describe the approach taken;<br> i.e. a modelling methodology which transcends the accepted<br> mathematical models for specific applications. This methodology<br> abstracts general models from first principles, by<br> employing an existing notation (bond graphs) as a metalanguage<br> for describing physical systems. This book is, therefore,<br> concerned with separating out the model development process<br> from the functions for which the model is developed, in order<br> to enhance understanding of the essentials of the real physical<br> systems. This book is organised in two parts, so that the reader may<br> first understand the motivation and the basic concepts, and<br> then have the proposed methodology illustrated by a variety of<br> examples covering a wide selection of applications. The first part describes general modelling principles, based on<br> system decomposition, first using classical dynamical analysis<br> and then via the energy bond graph notation. Bond graphs are<br> shown to provide a powerful core model representation from<br> which a variety of mathematical models may be derived. Bond<br> graphs provide a useful means of illustrating causality which<br> is shown to be a crucial aspect of system modelling. The second part uses specific case studies to illustrate the<br> application of this methodology to systematic generation of the<br> most widely used mathematical models. Reference is made to a<br> computer-aided modelling tool (MTT), which is a research modelling<br> toolbox which uses bond graphs to support the modelling of<br> dynamic processes. This textbook was originally published by Prentice-Hall in 1996. The rights were returned to the authors on 21st July, 2003.