Collision Free Navigation of a Multi-Robot Team for Intruder Interception (1809.02319v1)

Abstract: In this report, we propose a decentralised motion control algorithm for the mobile robots to intercept an intruder entering (k-intercepting) or escaping (e-intercepting) a protected region. In continuation, we propose a decentralized navigation strategy (dynamic-intercepting) for a multi-robot team known as predators to intercept the intruders or in the other words, preys, from escaping a siege ring which is created by the predators. A necessary and sufficient condition for the existence of a solution of this problem is obtained. Furthermore, we propose an intelligent game-based decision-making algorithm (IGD) for a fleet of mobile robots to maximize the probability of detection in a bounded region. We prove that the proposed decentralised cooperative and non-cooperative game-based decision-making algorithm enables each robot to make the best decision to choose the shortest path with minimum local information. Then we propose a leader-follower based collision-free navigation control method for a fleet of mobile robots to traverse an unknown cluttered environment where is occupied by multiple obstacles to trap a target. We prove that each individual team member is able to traverse safely in the region, which is cluttered by many obstacles with any shapes to trap the target while using the sensors in some indefinite switching points and not continuously, which leads to saving energy consumption and increasing the battery life of the robots consequently. And finally, we propose a novel navigation strategy for a unicycle mobile robot in a cluttered area with moving obstacles based on virtual field force algorithm. The mathematical proof of the navigation laws and the computer simulations are provided to confirm the validity, robustness, and reliability of the proposed methods.