A Dynamic-Model-Based Predictive Controller for a Novel Pendulum-Driven Spherical Robot

Abstract

Benefitted from its spherical shell, the spherical robot can work in hazardous environments and can easily return to normal after a disturbance. So, it may be a promising choice for missions such as planetary exploration, cave exploration, and so on. However, the internal structure of spherical robots can only provide a small space for the load, and the tightly coupled dynamical system of spherical robots makes it difficult to control. In this paper, a new internal structure is designed and a state-space representation of the dynamic model is given. In addition, the Model Predictive Control (MPC) based on the dynamics model was introduced so that the robot can track the desired trajectory without the need for decoupling. The dynamic model and the MPC controller were validated with different trajectories in simulations.