Flocking trajectory control under faulty leader: Energy-level based election of leader

A flocking algorithm using leader-follower strategy is developed, using energy, navigation control and radius of communication. The system consists of multiple agents having varying velocities and second-order dynamics. At any instant the group has one leader. All agents follow this leader, based on the observation of the position and velocity of the leader by a subset of the number of agents. Agents within the communication radius of other agents are able to establish connection and communicate with each other, either the position of the leader and/or the velocity of the leader. In this paper, we propose an energy-level based election of leader for flocking trajectory control under faulty leader. All agents compute certain energy. Agent that has the least energy is elected leader. A leader fault identification method has been developed, based on graph theory, applied to communication establishment. The leader is isolated as faulty if the agents within the communication neighbourhood of the leader are unable to interact with the leader. Under, these circumstances, these agents call for an election of leader, and the elected leader has the least energy amongst the rest. Regardless of isolation of the current leader and election of a new leader, the agents will converge to the desired trajectory.