Development of optimization toolbox with artificial intelligence methods

OData support
Dr. Harmati István
Department of Control Engineering and Information Technology

In the life of engineers, one of the usual problems is the optimization. It can be coding, performance measuring or infrastructure building, that is important the result should run as optimal process.

This thesis project introduce a few artificial intelligence based algorithms which created in the last two decades. These algorithms solve that optimization problem where they maximize or minimize their costfunction for global optimum.

All methods are looking for the global optimum, thus the function includes valleys then the methods will not return with local optimum. Another important feature that the methods harmonize their multiagent environment. Therefore these methods relate to the artificial intelligence because more agents/particles take part in the process and help each other to figure out the result.

This thesis project will show some swarm intelligence methods. These methods refer to the symptom/creature where their names come from. For example, the Bee algorithm imitates the pathfinding for nectar, tha Ant algorithm imitates the pheromon based pathfinding and the Bat algorithm imitates the hunting for preys. There are methods which not relate to animals but relate to human behavior as concept. The Basic Harmony Search shows how musicians play during a jam session.

This thesis will describe that the user can see in the MATLAB program/toolbox, which related to this document, how these methods work and solve problems in short time with small processor usage. This document introduces the reader into the process, the nature and the concept of the methods.


Please sign in to download the files of this thesis.