Date of Award

2008

Publication Type

Master Thesis

Degree Name

M.A.Sc.

Department

Electrical and Computer Engineering

Keywords

Engineering, Electronics and Electrical.

Supervisor

Kobti, Ziad (School of Computer Science), Kar, Narayan (Electrical & Computer Engineering)

Rights

info:eu-repo/semantics/openAccess

Abstract

This thesis presents a method of adapting IIR filters implemented as lattice structures using a Genetic Algorithm (GA), called ZGA. This method addresses some of the difficulties encountered with existing methods of adaptation, providing guaranteed filter stability and the ability to search multi-modal error surfaces. ZGA mainly focuses on convergence improvement in respects of crossover and mutation operators. Four kinds of crossover methods are used to scan as much as possible the potential solution area, only the best of them will be taken as ZGA crossover offspring. And ZGA mutation takes the best of three mutation results as final mutation offspring. Simulation results are presented, demonstrating the suitability of ZGA to the problem of IIR system identification and comparing with the results of Standard GA, Genitor and NGA.

Share

COinS