An ANN – Constant power generation control for LVRT of grid-connected PVG

This paper discusses an approach low-voltage ride-through control of a boost-converter for a two-stage grid-connected PV system under balanced grid voltage dips (BGVD). A combination of a maximum power point tracking (MPPT) and constant power generation (CPG) to control the boost-converter, respectively, in normal operation mode and BGVD under rapidly changing climatic conditions is proposed. MPPT and CPG commands based on artificial neural network algorithms are developed. Their goal is to maintain the operation of the low-voltage grid-connected photovoltaic system under instantaneously BGVD. This approach ensures a balance between the PV power generation and the injected active power when BGVD appears at the common coupling point (CCP). It permits to regulate the PV power to a specific value, to limit the over current at the CCP and to avoid energy aggregation at the DC-link bus. Simulations of 2.5 kW of a two-stage grid-connected PV system to verify the effectiveness of the command approach are presented.