An investigation of robotic swarm behaviour using decentalized control algorithms

Abstract

Although the roles of single and small multi-robot systems have been widely accepted and adopted for the needs of our society, the roles of swarm robotic systems are still under investigation. Swarm robotic systems typically mimic the behaviour of groups of social insects and animals. They can bring about a number of improvements over traditional multi-robot systems due to their adaptive nature. The simple individual behaviour of each agent within a swarm, can in turn result in complex collective swarm behaviour. Having a large number of robots equipped with basic sensing and processing capabilities and following basic rules makes it attractive for applications such as search and rescue or clean-up of toxic materials. The main aim of this thesis was to conduct a review of typical swarm behaviours and investigate the algorithms that are used to replicate such behaviour in robotic systems. After conducting the literature review, two algorithms were chosen and simulated to understand the emerging behaviour. The selected robot hardware was upgraded such that the algorithms could be implemented and demonstrated. This led to a better understanding of the topic of swarm robotics wherein the results were used to evaluate the degree of effectiveness of the chosen algorithms.