ANN-based voltage dip mitigation in power networks with distributed generation

Abstract

In the current power and energy scenario, distributed generation (DG) has generated a lot of interest across the globe due to the growing concerns about gradual depletion of fossil fuels, steep load growth, environmental pollution and global warming caused by greenhouse gas emissions. Renewable DGs such as wind generators and solar photovoltaic are well-recognized now-a-days as sources of clean energy. Voltage dips have always been a serious problem in electricity networks accounting for the disruption, poor power quality and affecting the cost and productivity of power utilities and the consumers. Therefore, with more DG penetration into the network, utilizing the DGs for improving power quality through voltage dip mitigation has become an important area of research in itself. In this context, this paper presents a novel technique for voltage dip mitigation with Distributed Generation (DG) using a simple feed forward Artificial Neural Network (ANN) to mitigate the effects of the power quality problems arising out of voltage dips in a distribution network. The scheme is simulated in DIgSILENT Power Factory 14.0 software and the tests are carried out on IEEE 9-bus test system. A three-phase short circuit fault on a line is simulated as the disturbance causing the voltage dip in the system. The model is trained, tested and validated in Matlab using mean square error and regression analysis.