Dynamic voltage restorer quality improvement analysis using particle swarm optimization and artificial neural networks for voltage sag mitigation

Abstract

Power quality is one of the problems in power systems, caused by increased nonlinear loads and short circuit faults. Short circuits often occur in power systems and generally cause voltage sags that can damage sensitive loads. Dynamic voltage restorer (DVR) is an efficient and flexible solution for overcoming voltage sag problems. The control system on the DVR plays an important role in improving the quality of voltage injection applied to the network. DVR control systems based on particle swarm optimization (PSO) and artificial neural networks (ANN) were proposed in this study to assess better controllers applied to DVRs. In this study, a simulation of voltage sag due to a 3-phase short-circuit fault was carried out based on a load of 70% of the total load and a fault location point of 75% of the feeder’s length. The simulation was carried out on the SB 02 Sibolga feeder. Modeling and simulation results are carried out with MATLAB-Simulink. The simulation results show that DVR-PSO and DVR-ANN successfully recover voltage sag by supplying voltage at each phase. Based on the results of the analysis shows that DVR-ANN outperforms DVR-PSO in quality and voltage injection into the network.