Neural network based fuzzy identification and its application to modeling and control of complex systems

This paper proposes a novel fuzzy identification approach based on an updated version of pi-sigma neural network. The proposed method has the following characteristics: 1) The consequence function of each fuzzy rule can be a nonlinear function, which makes it capable to deal with the nonlinear systems more efficiently. 2) Not only each parameter of the consequence functions but also the membership function of each fuzzy subset can be modified easily online. In this way, the fuzzy identification algorithm is greatly simplified and therefore is suitable for real-time applications. Simulation results show that the new method is effective in modeling and controlling of a large class of complex systems. >