Herding of a Flock of Agents Using Single Robot

In this paper, an algorithm is proposed for enabling a single robot to herd a flock of agents to a designated space or tracking a specified trajectory. The herding algorithm is designed using a second-order dynamic model of flocking based on Reynolds rules. At first we analyze the stability of the flock with or without external effect (herder) and conclude the conditions of avoiding fragmentation of the flock. From these conditions a herding algorithm is proposed, and it is also demonstrated using numerical simulations.