Optimal Static Var Compensator Switching in Unbalanced Distribution System Based on Artificial Neural Network

The rising incidence of equipment mal-operations and failures due to inadequate reactive power compensation facilities at power distribution centers is an increasing concern everywhere. The distribution static compensator (DSTATCOM) is the best solution for this issue. However, they can be very costly and complex to use. In this case study, the static var compensator (SVC) is used in a thyristor binary switched capacitors (TBSC)-thyristor-controlled reactor (TCR) configuration. The TCR in SVC, while correcting, produce harmonics intrinsically due to non-sinusoidal currents. The purpose of this study is to investigate the connection between unbalanced load with poor reactive power management and harmonic mitigation facility in a power distribution network. The optimal size reactor and the binary sequence of the capacitor banks are considered. The artificial neural network (ANN) approach is used to solve a nonlinear problem and calculate the optimal firing delay angle used to switch thyristors in TCR with accurate delay to reduce harmonic distortion significantly. The proposed strategy is appealing and can be used for the steady-state reactive power allocation in SVC applications in distribution networks.