Trajectory tracking control of a three-wheeled omnidirectional mobile robot using disturbance estimation compensator by RBF neural network

This paper presents a trajectory tracking approach of a three-wheeled omnidirectional mobile robot using disturbance estimation compensator by RBF (radial basis function) neural network. Model uncertainty, nonlinear terms in the dynamic model and unknown disturbances are treated as a ‘total disturbance,’ and estimated and compensated it by RBF neural network. The proposed controller is a modification of the PD CT (proportional derivative computed torque) method, and has a neural network compensation term that estimates and compensates for the ‘total disturbance.’ The effectiveness of the proposed approach is compared with PD CT, PID CT (proportional integrated derivative CT) controllers through simulations and experiments. Simulations and experimental results show that the proposed approach is more effective than the CT trajectory tracking approaches in the presence of model uncertainties and unknown disturbances. In particular, the proposed method is much more efficient in the presence of unknown disturbances as well as model uncertainties.