Predictive control for the navigation of spherical robots in obstacle-rich environments

This paper introduces an advanced predictive control algorithm tailored for spherical robots operating within obstacle-dense environments. The proposed strategy employs a sophisticated system model to forecast the robot's behavior over a defined future time horizon, capitalizing on the strengths of predictive control, which include accelerated convergence time, particularly beneficial in dynamic scenarios. We conduct a comprehensive comparative analysis between the predictive control approach and feedback linearization control, focusing on the robot's navigation through obstacles. Our findings indicate that the predictive control framework considerably improves the robot's overall motion performance, leading to shorter response times, enhanced convergence capabilities, and greater resilience when navigating obstacle-dense environments. Furthermore, this approach effectively minimizes control errors, achieving rapid convergence to zero. This research highlights the effectiveness of predictive control in optimizing the agility and accuracy of spherical robots in challenging operational settings. The proposed method represents the first algorithm capable of effectively addressing a comprehensive range of motion tasks for spherical robots. This includes trajectory tracking and both static and dynamic obstacle avoidance with global stability and optimal performance, which are essential for their development.