Application of a genetic algorithm-based PI controller in a spherical robot

The spherical robot is a nonholonomic system and takes two motors as impetus sources. It can realize any omnidirectional movement by changing the position of the barycenter of weights. But the existing spherical robot is controlled by remote control, so it doesn´t achieve the self-control in a real sense. In this paper, on the basis of the kinematic and dynamic models of the spherical robot, we start from the mathematic model of the DC servo motor. Subsequently, the mathematical model of the control system is constructed. PI controller is adopted to achieve the real-time control of the whole system. Then, genetic algorithm (GA) is applied in PI parameters setting and optimization to improve the stability of this system. In the end, the validity is verified by the simulation experiments.