Abstract
In this paper, we propose a novel control approach for nonholonomic spherical mobile robot using a network of neural oscillators. Key features of a neural oscillator include its ability to generate stable rhythmic signals, and to tune their frequencies and patterns depending on the environment. Motivated by this versatile entrainment property, we adopt the network as a controller for the spherical mobile robot with highly complicated controllability structure. We examined effectiveness and robustness of the proposed method by several numerical simulations.
