Swarm Algorithm Target Seeking In Unknown Environment

Goh, Shu Wan (2018) Swarm Algorithm Target Seeking In Unknown Environment. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Goh Shu Wan - Full Thesis.pdf Restricted to Registered users only Download (4MB)

Abstract

Swarm robot mean that a large group of robots work together and moving together. The criteria of the swarm robot is a group of robots that only have the sensing and actuating capabilities but no knowledge about the overall situation. In this research, the main purpose is to develop swarm target seeking algorithm in unknown environment called E-puck Swarm Algorithm (ESA). This research is focus of the development on swarm intelligent reacting to unknown environment such as jungle. Webots simulation software will be used to design the experimental scenarios to test the swarm ability. E-puck will be the swarm agents. A swarm of E-puck will added into the designed environment scenarios of Webots simulation software in order to test the developed swarm algorithm. The objective of this ESA is to search a target in the unknown environment and form a communication network if found target. Other than that, this swarm e-puck must have the abilities to avoid the obstacles to avoid damage on e-puck, able to maintain the secure distant with neighbouring e-pucks to avoid the collision and expanded the searching range because secure distant is to spread the e-puck instead of stick together, detect goal ability, able to pass the information among each other. ESA is the new concept based on the intelligent of swarm robot and rescue robot. There are a various experiment scenario conducted. They are conducted in different size of seeking area such as 1x1, 2x2, 3x3 and ,4x4 and different number of e-puck. The time to detect the presence of target, time to confirmed target and the time to formed the network were recorded in the table. However, the limitation in this research is the swarm robot is not able to back to the starting points. Navigation and localization of the swarm can be improved in the future. This algorithm is able to perform the target seeking, obstacle avoidance, avoid collision with other agents, form the communication network when found target.