Application of artificial intelligence techniques for optimizing the performance of solar photovoltaic systems

Abstract

Solar Photovoltaic (PV) systems indeed serve a significant role in the newlineevolution of renewable sources based electric power generation. The expansion of newlinesolar PV systems in recent years has been phenomenal and has emerged as one of newlinethe leading renewable energy technologies. Despite the substantial development newlineand progress of solar PV technology, there are still several concerns associated newlinewith its implementation and widespread adoption. The presence of sunshine in a newlineparticular area is contingent upon weather patterns, daily cycles, and seasonal newlinevariations. This intermittency of solar PV energy poses challenges for grid newlineintegration and requires efficient energy storage or backup power systems to newlineensure a consistent electricity supply. newlinePartially shading is one of the main concerns in achieving optimal PV newlinesystem performance. A PV panel that is partially shadowed receives significantly newlineless sunlight in comparison to the unshaded portion. It can lead to PV panel output newlinecurrent and voltage imbalances, which would impair both its general functionality newlineand power production. Some mitigating methods for partial shading include newlinebypass diodes, Maximum Power Point Tracking (MPPT) algorithms, PV array newlinereconfiguration methods, module-level power electronics such as micro-inverters newlineor power optimizers, and intelligent shading control techniques. In this work a newlineMPPT and PV array reconfiguration are employed to reduce the power losses newlineand improve performance and reliability of PV system under partial shading. newlineA methodology has been proposed a magic square array layout and differential newlineevolution-based adaptive PandO MPPT approach to reduce partial shading impacts newlineand increase PV system power production independent of shading patterns. This newlinestudy also investigates the robustness and scalability of the proposed method newlineacross different shading scenarios and PV array sizes newline