The application of a neural-fuzzy logic controller to process control

A neural-fuzzy controller is an intelligent system that allows for the combination of qualitative knowledge in fuzzy rules and the learning capabilities of neural networks. This paper examines the suitability of one particular neural-fuzzy model, the adaptive network fuzzy interference system (ANFIS) proposed by J.-S.R. Jang, for use as part of control systems. The adaptive neural-fuzzy controller developed uses the sign of the output error and the gradient descent algorithm to update its parameters and gives superior control than PID. The controller accommodates modifications made to the original plant.<
>