Connectivity constrained multirobot coordination for rendezvous and navigation

Abstract

This thesis aims to develop a new multirobot coordination control scheme with particular interest in preserving network connectivity in the presence of obstacles under bounded control input. This research consists of the following three elements. First, the modeling of a connectivity constrained multirobot system in the presence of obstacles is investigated. The connectivity constraint is carefully modeled in building a geometrical relationship amongst connected robots in disk-like communication mode. Obstacle avoidance, as an important constraint, is incorporated into the modeling of the geometrical relationship between robots and obstacles. These constraints are well considered in the multirobot control objectives, such as rendezvous and navigation, in this study. Second, a novel controller with bounded control input is developed to solve rendezvous tasks while preserving network connectivity in the presence of obstacles. Unlike existing approaches, where the rendezvous task is usually accomplished in an obstacle-free environment, the proposed method addresses the rendezvous control problem in the presence of obstacles, while maintaining connectivity. It is demonstrated that when the robots are connected in their initial configurations, the proposed controller enables the underlying network to remain connected during the whole motion evolution. Simulations and experiments are performed to demonstrate the effectiveness of the proposed controller. Third, a new approach to solving the multirobot navigation control problem while maintaining network connectivity in the presence of obstacles is further developed. With the proposed controller, the robot can approach each desired position while maintaining network connectivity and avoiding obstacles, provided that the initial robot configurations are connected and the desired configurations are attainable. Unlike existing approaches, this method addresses the network connectivity problem, while considering obstacle avoidance during multirobot navigation. This thesis makes an important contribution to the robotics community in that the connectivity constrained multirobot coordination scheme, with obstacle avoidance and bounded control input, is developed specifically to solve multirobot rendezvous and navigation problems.

Date of Award	15 Feb 2012
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Dong SUN (Supervisor)