Parameter Estimation of Neuro-Fuzzy Model by Parallel and Series-Parallel Identification Configurations

In this paper combinations of two well-known identification methods namely series-parallel and parallel configurations, are proposed to identify the learning parameters of neuro-fuzzy inference system. Two new configurations out of four possible combinations in identifying parameter of the neuro-fuzzy system are proposed. These two proposed configurations are devised by applying series-parallel configuration to premise part and parallel configuration to consequent part of neuro-fuzzy system and vice versa. The proposed configurations have been compared with already existing, namely series-parallel and parallel configurations (to both premise and consequent part of neuro-fuuzy system) on two non-linear dynamic systems.