A Hybrid, Multi-Agent Model of Foraging Bottlenose Dolphins
Haque, Musad A.
Rahmani, Amir R.
Egerstedt, Magnus B.
MetadataShow full item record
Social behavior of animals can offer solution models for missions involving a large number of heterogeneous vehicles, such as light combat ships, unmanned aerial vehicles, and unmanned underwater vehicles. We draw inspiration from the foraging techniques of bottlenose dolphins to address the problem of heterogeneous multi-agent herding. We produce a hybrid automaton model of the entire foraging method - search, detect, and capture - where agents are modeled as first-order systems in which interactions are defined through spatial proximity. Finally, simulations are provided to illustrate that our model is expressive enough to capture this complex biological phenomenon.