Development of a predicitive type-2 neurofuzzy controller

Abstract

A controller that combines the main characteristics and advantages of three different control methodologies is proposed for the control of systems with nonlinearities and uncertainties. A neural network predictive control approach is implemented modifying the output of a controller with a fuzzy logic structure that uses type-2 fuzzy sets. Neural networks are also used to optimize the membership function parameters. The proposed controller is tested by simulation for the control of a bioreactor characterized by bifurcation and parameter uncertainty.