Modified continuous ant colony algorithm for function optimization

Authors

Alexandre Aidov, Florida International University

Document Type

Thesis

Degree

Master of Science (MS)

Major/Program

Mechanical Engineering

First Advisor's Name

George S. Dulikravich

First Advisor's Committee Title

Committee Chair

Second Advisor's Name

Yiding Cao

Third Advisor's Name

Igor Tsukanov

Date of Defense

7-15-2008

Abstract

Many classical as well as modern optimization techniques exist. One such modern method belonging to the field of swarm intelligence is termed ant colony optimization. This relatively new concept in optimization involves the use of artificial ants and is based on real ant behavior inspired by the way ants search for food. In this thesis, a novel ant colony optimization technique for continuous domains was developed. The goal was to provide improvements in computing time and robustness when compared to other optimization algorithms. Optimization function spaces can have extreme topologies and are therefore difficult to optimize. The proposed method effectively searched the domain and solved difficult single-objective optimization problems. The developed algorithm was run for numerous classic test cases for both single and multi-objective problems. The results demonstrate that the method is robust, stable, and that the number of objective function evaluations is comparable to other optimization algorithms.

Identifier

FI13101551

Recommended Citation

Aidov, Alexandre, "Modified continuous ant colony algorithm for function optimization" (2008). FIU Electronic Theses and Dissertations. 1166.
https://digitalcommons.fiu.edu/etd/1166