Indirect adaptive control of a robot using fuzzy neural network

The paper demonstrates that the fuzzy neural network can be used effectively for the control of an actual industrial robot arm. The control scheme consists of a multi-layer feed-forward neural network to model the plant and a fuzzy neural network is used as a controller. The proposed design will recognise changes in process dynamic and adjust its parameters rapidly to counteract for these changes. Experimental results show that the incorporation of fuzzy logic into the network structure can provide a priori information and thereby speed up the network convergence.