Resource management with kernel-based approaches for grid-connected solar photovoltaic systems

Abstract

The increasing penetration of photovoltaic (PV) power generation into the distribution grids has resulted in frequent reverse active power flows, rapid fluctuations in voltage magnitudes, and power loss. To overcome these challenges, this paper identifies the resource management of grid-connected PV systems with active and reactive power injection capabilities using smart inverters. This approach is aimed to minimize the voltage deviations and power losses in the grid-connected systems to accommodate the high penetration of PV systems. A kernel-based approach is proposed to learn policies and evaluate the reactive power injections with smart inverters for improving grid profile, minimizing power losses, and maintaining safe operating voltage limits. The proposed approach performs inverter coordination through nonlinear control policies using anticipated scenarios for load and generation. To assess the performance of the proposed approach, numerical simulations are performed with a single-phase grid-connected PV system connected to an IEEE bus system. The results show the effectiveness of the proposed approach in minimizing power losses and achieving a good voltage regulation.