Neural Network Based Control of Dual Mode VSI in HRESBS Under Abnormal Grid Scenario

The main focus of this work is on the performance of the hybrid renewable energy source-based system (HRESBS) under distorted grid and imbalanced nonlinear load conditions. The system's objective is to supply consistent power to DC charging stations and AC local loads under dynamic conditions while maintaining overall power quality. A highly selective filter (HSF) is used in grid interfaced mode (GIM) to suppress the harmonics component and imbalance present in grid voltages in the event of abnormal grid conditions. Moreover, utilizing HSF as a pre filter to a second order generalized integrator frequency locked loop (SOGI-FLL), the fundamental component of load current under the effect of noise and DC offset is determined accurately. Additionally, in off-grid or autonomous mode (AM), interfacing voltage source inverter (VSI) control is carried out using artificial neural network (ANN) based control to generate precise reference three leg inverter currents under dynamics. The presented topology is simulated in MATLAB Simulink platform under various perturbations to obtain the effective response of the system in both GIM and AM operating conditions.