Recurrent Fuzzy Cerebellar Model Articulation Neural Network Based Power Control of a Single-Stage Three-Phase Grid-Connected Photovoltaic System during Grid Faults

A recurrent fuzzy cerebellar model articulation neural network (RFCMANN)-based controller to regulate the active and reactive power of a single-stage three-phase grid-connected photovoltaic (PV) system during grid faults is proposed in this study. Since the rapid growth of the amount of the PV systems in recent year has greatly affected the stability of the power system, the grid codes demand the grid-connected PV systems to have the low-voltage ride through (LVRT) capability to help sustaining the stability of power system especially during the grid faults. To satisfy the LVRT requirements and ensure the injected current within the safety value, the active and reactive power commands are calculated by using the current profile of the LVRT grid requirements and the current limit of the inverter. Moreover, the proposed RFCMANN controller uses the signed distance and input space repartition mechanisms to convert the dual-input variables to sole-input variable and repartition the input space to an appropriate quantity. Therefore, the structure and computation complexities of the proposed RFCMANN controller are effectively reduced and make it more practical. Furthermore, varied learning-rate coefficients are designed to guarantee the convergence of the tracking error for the proposed RFCMANN controller.