A multi-robot cooperative hunting approach based on dynamic prediction of target motion

Abstract

Multi-robot hunting receives much attention due to its natural antagonism. In this paper, a hunting approach based on dynamic prediction of target motion is proposed where the prediction step is optimized according to current environment. Based on the predicted target positions, encirclement points corresponding to different prediction steps are obtained. Then, an optimized prediction step is determined according to the variance of the distances between the robots and their respective desired encirclement points. Finally, the robots shrink to capture the target. The validity of the proposed approach is verified by simulations.