Enhancing Spherical Rolling Robot Control for Slippery Terrain

Abstract

Spherical rolling robots represent an innovative frontier in robotics, characterized by advanced designs and sophisticated internal mechanisms. This study introduces a spherical robot that leverages the conservation of angular momentum through the use of rotating flywheels, enabling it to execute complex trajectories even on slippery terrain. In contrast to conventional approaches, that often render the system nonholonomic by combining internal driving mechanics with control models involving rolling without slipping conditions, this article demonstrates the feasibility of controlling a robot capable of generating omnidirectional torques and maintaining robust control even in the presence of unexpected slipping. This is made possible through the use of a nonlinear control algorithm called Feedback Local Optimization Principle (FLOP), which has the advantage of incorporating into the modeling nonlinear effects such as contact friction force. The results highlight that the sphere controlled based on a dynamic model with slipping can achieve superior performance and reach the target more effectively compared to traditional modeling approaches that do not account for slipping.