Identification of dynamic systems using recurrent fuzzy neural network

This study proposes a recurrent fuzzy neural network (RFNN) structure, which is a modified version of a fuzzy neural network (FNN). The proposed RFNN is a recurrent multilayered connective network for realizing the fuzzy inference and can be constructed from a set of fuzzy rules. Adding feedback connections in the second layer of the FNN develops the temporal relations embedded in the RFNN. This modification provides the memory elements of the RFNN and expands the basic ability of the FNN to include temporal problems. Since a recurrent neuron has an internal feedback loop, it captures the dynamic response of a system, thus the network model can be simplified. Finally, the proposed RFNN is applied to identify some nonlinear dynamic systems. Simulation results confirm the effectiveness of the RFNN.