Improving the performance of cooperative search by using communication and coordinaton in multi robot systems

In this thesis the effects of communication on cooperative robotic search task is studied through both simulations and robotic experiments. Firstly, the benefits of networked communication are investigated by evaluating the results of communicative and non-communicative search by multi robot teams in a directional area without any obstacles. In the continuing studies the robotic search experiments are repeated with more robots and in a more realistic area containing walls and obstacles. During the search each robot is also allowed to keep an occupancy grid based map of the environment containing the information about the cells it has visited and to exchange this information with the other robots within its communication range. The effect of the size of communication range on the performance of the system defined as the time of completion the cooperative search task was investigated through both simulations and experiments. In the later part a robotic dispersion task which is a preliminary work for a robotic search is studied. During the dispersion of the robots any other positioning technique rather than the received radio signal strengths broadcasted from the other robots is not used. The dispersion of the robots is provided by using bacterial foraging algorithm which is an optimization method inspired from the foraging behaviors of mono-cellular microorganisms.