Design and Simulation of Dynamic Voltage Restorer based on Fuzzy Controller Optimized by ANFIS

Abstract

The fuzzy logic controller (FLC) appears to be the unique solution when the process is too complex for analysis by conventional techniques or when the available information data are interpreted qualitatively, inexactly or with uncertainty. In literature, the proposed FLC in general consists of two inputs (error and derivative of error) and one output. The number of membership functions is chosen in most cases to be five or seven regardless of the approach used for the design. In this paper, we propose Adaptive Neuro-Fuzzy Inference System (ANFIS) approach to optimize the two inputs one output FLC with seven membership functions to one input one output FLC with three membership functions without compromising accuracy. The study is applied to control a Dynamic Voltage Restorer (DVR) in voltage sag/swell mitigation. The results of simulation using MATLAB/SIMULINK show that the performance of the optimal FLC generated by ANFIS is comparable with the initial given FLC.DOI: http://dx.doi.org/10.11591/ijpeds.v4i2.5685