Simulation application and comparison of swarm based optimization algorithms on source localization problem of swarm robotics

In this thesis, we applied swarm-based optimization algorithms to problem of finding source location of swarm robotics, and we suggested new adaptations to those algorithms if needed, then we developed a simulation program which can simulate those algorithms. We analyzed success of source tracking and source localization of swarm robotics in, single source localization, multiple source localization which have homogenious signal intensitiy and heterogenious signal intensity, and dynamic source localization scenarios. We analyzed and compared swarm-based optimization algorithms such as particle swarm optimization, artificial bee colony, quick artificial bee colony and modified bee colony. We also suggested and integrated modification to those algorithms in order to apply those solutions in real world applications. With suggested modifications we developed a source localization in swarm robotics simulation application with C# language in UNITY 3D. We developed same application in MATLAB in order to compare metaheuristic methods more comprehensible with figure libraries of MATLAB. We analyzed non-numerical matching problem, in order to get better sense of swarm-based optimization algorithms' skill set. We suggested artificial bee colony algorithm and developed an application with MATLAB for matching solution of Chinese postman problem.