Adaptive multi-model controller for robotic manipulators based on CMAC neural networks

In this paper, an adaptive multi-model controller based on CMAC neural networks (AMNNC) is developed for uncertain nonlinear MIMO systems. AMNNC is a kind of adaptive feedback linearizing controller where nonlinearity terms are approximated with multiple neural networks. The weighted sum of the multiple neural networks is used to approximate the system nonlinearity for a given task. The proposed control scheme is applied to control a robotic manipulator, where some varying tasks are repeated but information on the load is not defined; it is unknown and varying. It is shown how the proposed controller is effective because of its capability to memorize the control skill for each task using neural networks. Simulation results demonstrate the effectiveness of the proposed control scheme for the robotic manipulator, in comparison with the conventional adaptive neural network controllers (ANNC)