Abstract:

Patrolling is the act of traveling around an area at regular intervals in order to provide security or supervise its current state. Multi-agent patrolling has received growing attention in recent years because of its wide range of potential applications. The main objective of this thesis is to develop an autonomous patrolling system composed of a multi-robot team to achieve efficient patrolling. The proposed system can derive patrolling routes given a territory and specific locations that need to be patrolled in the territory, and the patrolling routes can be adjusted according to different patrolling requirements in order to achieve the best patrolling performance. In addition, the system will check whether partitioning a territory will improve patrolling performance or not. If the answer is affirmative, then optimization techniques will be employed to determine the best partition and derive the optimal allocation for the team of robots. These robots will repeatedly patrol their each local task locations until an emergency occurs. Once an emergency arises, the system will adjust the assigned patrolling routes of the rest of active robots to assist other robots to enhance the robustness of the system. Finally, the simulation results and the physical experiments show that the proposed system can accomplish patrolling tasks autonomously and cooperatively.