Particle swarm optimization — Based identification of a bouncing spherical robot

In this paper a spherical robot consists of a spherical steel spokes shell and an inner bouncing mechanism. To control the bouncing mechanism, a system model should be introduced. The system model only deal with the first bounce of the spherical mobile robot to overcome the bouncing restitution for landing. This paper presents a transfer function model prediction with particle swarm optimization for a bouncing spherical mobile robot. The particle swarm optimization technique is used in search for accurate model with capability to adapt with different input height of drop. The model is further validated with autocorrelation and cross-correlation test and it is proven to give an error tolerance between the 95% confidence limit.