A hybrid approach for trajectory tracking control of redundant robot manipulators

In this paper, a hybrid approach is presented for tracking control of redundant robot manipulators. In this approach, an RBF(Radial Basis Function) neural controller and an adaptive bound is added to the model based controller i.e (CT type controller). The hybrid controller achieves end-effector tracking as well as subtask tracking adequately. The RBF network learns all the existing uncertainties in the system and requires no offline learning. The adaptive bound part approximates unknown bounds on unstructured uncertainties as well as neural network reconstruction error effectively. The Lyapunov approach is utilized to prove the closed loop system to be asymptotic stable. To validate the effectiveness of control scheme numerical simulation study is performed.