Candidate: Saad Y. Yasin

Degree of: Doctor of Philosophy

Department: Mechanical and Aeronautical Engineering

Title: Systematic Methods for Designing Fuzzy Logic Controllers with Theoretical Foundations and Practical Applications

Committee:
Dr. R. P. Sharma, Chair
Dr. J. H. Hamelink
Dr. S. Ganesan
Dr. K. Naghshineh
Dr. G. S. Alag

Date: Friday, March 1, 2002, 3:00 p.m. - 5:00 p.m.
3014 Kohrman Hall

Abstract:
Two newly developed models for formulating analytical fuzzy control laws, the left and right parameterization and normalized spline-base membership function, were studied. The developed models were then used in two important and practical applications. The study showed that a fuzzy logic controller based on the developed models, if developed, would effectively control highly nonlinear and complex systems. This study also investigated the application of fuzzy control in areas not fully utilizing fuzzy logic control. Three important practical applications pertaining to the automotive industries were studied. In the first application, the idle speed of spark ignition engines, two types of fuzzy controllers were developed. A conventional controller-like performance fuzzy controller was obtained by using the two methods and intuitive knowledge of the idle speed control system. In the second application, the automotive cruise control problem, an adaptive Proportional plus Integral plus Derivative (PID)-type fuzzy logic controller was developed. Results that were similar to, and in certain cases better than, those obtained using linearized conventional PID and linear quadratic regulator (LQR) controllers were shown. In the third application, the air/fuel ratio control problem, fuzzy clustering techniques and experimental data were utilized to develop an air/fuel ratio fuzzy controller.
Comparison with recently published conventional control-based studies showed a similar control ability of the developed controller. The influence of the fuzzy inference operators and parameters on performance and stability of the fuzzy logic controller was studied. Results indicated a great effect of certain parameters or combinations on the performance and stability of the fuzzy controller. Diagnostic guidelines to identify inadequate performance were developed.

Related Topics

Main List of Archives:
Dissertation Defenses

Current Dissertation Defenses