Abstract
We introduce a control logic which allows a group of autonomous ‘pursuers’ to herd a single, non-hostile ‘evader’ along a desired trajectory. They achieve this with a general knowledge of the repulsion scheme on the evader and utilizing a feedback control, based on the evader state and the known desired trajectory of the evader. The quasi-decentralized control of the pursuers is a version of a region holding sliding mode controller (SMC). The novelty of this work lies in the decentralized distribution of pursuers within a desired region whose boundaries are maintained utilizing the robust properties of SMC. The dynamic pursuer distribution repels the evader with a desired force which reduces its tracking error. Illustrative examples are studied for validation.
Get full access to this article
View all access options for this article.
