Masters Theses
Date of Award
8-1997
Degree Type
Thesis
Degree Name
Master of Science
Major
Electrical Engineering
Major Professor
J. D. Birdwell
Committee Members
Jack S. Lawler, Tse-Wei Wang
Abstract
This research presents a new approach to process modeling which attempts to capture the dynamic behavior of object-oriented signals. The new modeling approach, finite impulse response hybrid system modeling, is a method that combines the theories of Markov chains and finite impulse response (FIR) filters. The finite impulse response hybrid system can be used to represent a special class of dynamic processes where relationships among process variables are poorly understood and hidden inside a large quantity of available data. Traditional mathematical methods of modeling do not offer the flexibility provided by the new approach since no assumptions of linearity or other restrictions of process behavior are made. A digital modeling and simulation package, implemented using Matlab is developed. The effectiveness of finite impulse response Markov chain has been demonstrated by its application to modeling a highly nonlinear exothermic reaction process. The performance of the model accurately predicts the future output events of a process based on a set of past input events.
The new modeling approach is best suited for processes which generate large and messy data in which clear cause-effect relationships are not always clearly observed or understood. The model thus developed has a high potential for it to be used in subsequent process control applications, to allow for tighter control possible than that of conventional modeling approach.
Recommended Citation
Ong, Sze Wei, "Finite impulse response hybrid system modeling. " Master's Thesis, University of Tennessee, 1997.
https://trace.tennessee.edu/utk_gradthes/10671