Adaptive neuro-fuzzy controller based on simplified ANFIS network

A novel technique used to design a simple version of an adaptive neuro-fuzzy controller (ANFC) is described in this paper. The structure of the proposed Adaptive Simplified Neuro-Fuzzy Controller (ASNFC) consists of reduced numbers of input membership functions (MFs) and consequence parameters (CPs). A Neuro Identifier (NI) is used to track the behaviour of the plant on-line and update the controller. The ASNFC is applied to an SVC device, located at the middle of a single machine infinite bus system (SMIB), to damp power system oscillations. Results of simulation studies demonstrate that the proposed ASNFC provides similar control actions as the ANFC, but with less parameters to optimize. Although the proposed controller is a simplified version of the ANFC, the simulation results obtained show system responses similar to that with the ANFC.