Supervisory recurrent fuzzy neural network control for long-term ecological systems

Abstract

This paper develops an intelligent ecological biomass management method called supervisory recurrent fuzzy neural network control (SRFNNC) to deal with the long-term management of ecological system, which is an uncertain nonlinear system subject to unpredictable but bounded disturbances. This SRFNNC system is composed of a recurrent fuzzy neural network (RFNN) controller and a supervisory controller. The RFNN controller is investigated to mimic an ideal controller and the supervisory controller is designed to compensate for the approximation error between the RFNN controller and the ideal controller. This SRFNNC is employed to keep the biomasses of an ecological system within a small neighborhood of the unique nontrivial optimal equilibrium state of the undisturbed ecosystem. By applying this controller, the accumulative yield of harvest is better than that obtained with state feedback control and no control.