Shepherding algorithm for heterogeneous flock with model-based discrimination

The problem of guiding a flock agents to destination by the repulsion forces exerted smaller number external is called shepherding problem. This has attracted attention due its potential applications, including diverting birds away for preventing airplane accidents, recovering spilled oil in ocean, and swarm robots mapping. Although there have been various studies on problem, most them place uniformity assumption dynamics be guided. However, we can find practical situations where this does not necessarily hold. In paper, consider situation which are given consisting normal guided other variant agents. Under situation, propose method sheep. Specifically, method, shepherd discriminates based their behaviors' deviation from one predicted potentially inaccurate model As discrimination process, two methods using static dynamic thresholds. Our simulation results show that proposed outperform conventional types