Find in Library

Navigation and trajectory planning in environments with background flow, relevant for robotics, are challenging provided information only on local surrounding. The authors propose a reinforcement learning approach for time-efficient navigation of a swimmer through unsteady two-dimensional flows.