Electronic Thesis and Dissertation Repository

Degree

Master of Engineering Science

Program

Civil and Environmental Engineering

Supervisor

Dr. Slobodan Simonovic

Abstract

The Differential Evolution (DE) algorithm is a powerful search technique for solving global optimization problems over continuous space. The search initialization for this algorithm does not adequately capture vague preliminary knowledge from the problem domain. This thesis proposes a novel Fuzzy Differential Evolution (FDE) algorithm, as an alternative approach, where the vague information of the search space can be represented and used to deliver a more efficient search. The proposed FDE algorithm utilizes fuzzy set theory concepts to modify the traditional DE algorithm search initialization and mutation components. FDE, alongside other key DE features, is implemented in a convenient decision support system software package. Four benchmark functions are used to demonstrate performance of the new FDE and its practical utility. Additionally, the application of the algorithm is illustrated through a water management case study problem. The new algorithm shows faster convergence for most of the benchmark functions.

DEO-Code.zip (2657 kB)
DEO Software Code

DEO-Install.zip (1769 kB)
DEO Installation Software Files