Efficiency of Swarm-Intelligence Methods in Engineering Applications

OData support
Supervisor:
Kertész Zsolt László
Department of Control Engineering and Information Technology

Computational Intelligence has passed through a significant evolution in the past two decades and bio-inspired approaches have also become increasingly popular. The need for efficient algorithms from various fields of engineering inspiring researchers to develop new optimization techniques and algorithms. Swarm intelligence is a modern bio-inspired artificial intelligence discipline which is deeply embedded in biological study of self organizing social insects. These biological systems are robust, flexible and scalable, which advantages make swarm intelligence based algorithms and multi agent systems more successful over the traditional design paradigms. There are lots of optimization algorithms available now and we have to investigate, where can we use these techniques and how can we make profit from these bio-inspired techniques. In the paper I focused on three of the most successful optimization techniques based on swarm intelligence: Ant Colony Optimization, Particle Swarm Optimization and Artificial Bee Colony Optimization, then examining the engineering areas have been using them in the view of their efficiency and limits. I furthermore focus on a multi agent programming and visualization environment software called NetLogo, by the Center for Connected Learning (CCL) Institute. NetLogo is a perfect tool for visualize complex multi agent systems, specifically swarm intelligence based ones. Swarm intelligence is a novel approach to multi agent systems design, for example collective robotics. In the paper I present a brief review of Swarm Robotics and two noticeable multi-robot projects: SYMBRION-REPLICATOR and Swarm-bots.

Downloads

Please sign in to download the files of this thesis.