Towards formalism of deep model descriptions of dynamic systems
The deep models which is referred to as the structural and behavioral models have been used in system analysis of dynamic systems. The description of such models, however, is far away from being formalized. This paper is devoted to the formalized descriptions of deep models of dynamic systems. In this paper, some basic concepts such as module, structure and hierarchical decomposition model are defined first and the related problems such as the automatic generation of the behavioral models of the higher level components from the behavioral models of the components below are then discussed with the premise that the system considered is linear. The problem about how to represent the model knowledge in knowledge-based systems is also discussed. Finally, a simulation example of applying deep models to fault diagnosis of a dynamic system is presented with the intention not only to illustrate the applications of deep models but also to present an approach for fault diagnosis of dynamic systems.
Unable to display preview. Download preview PDF.
- DAVIS,R. 1983. “Reasoning From First Principle in Electronic Troubleshooting.” Internat. J. Man-Mach. Stud. 19, 403–423.Google Scholar
- DAVIS,R. 1984. “Diagnostic Reasoning Based on Structure and Behavior.” Artif. Intell. 24, 347–410.Google Scholar
- LEARY,J. J. and GAWTHROP, P. J. 1987. “Process Fault Detection Using Constraint Suspension.” IEE PROCEEDINGS Pt.D. 134(4), 264–271.Google Scholar
- YUAN,Z. G. and WEN,C. Y. 1989 “Using Expert Systems in Failure Diagnosis of Dynamic Systems.” In Proceedings of 1989 Beijing International Conference on System Simulation and Scientific Computing(Beijing, Oct.1989), 751–756.Google Scholar
- YUAN,Z. G. and WEN,C. Y. 1990. “Structural Analysis Method for Failure Diagnosis in Dynamic Systems.” In Proceedings of 1990 Summer Computer Simulation Conference(Calgery, July 1990) 319–324.Google Scholar