Self-organizing neural fuzzy inference network for intelligent control

A self-organizing neural fuzzy inference network (SONFIN) is proposed. This network is built for realizing the basic elements and functions of dynamic fuzzy inference, considered like a series of dynamic fuzzy rules. Each node and weight represent an element in a fuzzy rule. These are created online based an a concurrent structure and parameter identification. The structure learning together with the parameter learning forms the learning algorithms for a dynamic neural fuzzy network. The aligned clustering-based algorithm is used for reducing the number of generated fuzzy rules. Efficiency of the SONFIN is verified from simulation results.