Comparative analysis of Dynamic Voltage Restorer based on PI and ANN Control strategies in order to improve the voltage quality under Non-linear loads.

Abstract

Voltage sag, swell and harmonic distortion are the key power quality concerns addressed by the distribution network. The presence of non-linear loads and renewable energy sources like solar, wind etc. is the reason for power quality issues. These necessitates the need for the development of a power quality conditioner to compensate the effects of these power quality problems. Hence a Dynamic Voltage Restorer is developed and deployed to improve power quality by decreasing harmonics and adjusting for voltage swell and sag. DVR control is achieved by regulating the load voltage under a variety of unexpected working conditions. Artificial Neural Network controller based DVR is carried out. The simulation results show that the Artificial Neural Networks based DVR with a THD of 2.84 % outperforms the PI-based controller THD of 4.48%.