Model-based strategies for fault detection and diagnosis in grid connected photovoltaic plants

The need of sustainable energy solutions is a demand nowadays due to the high electricity consumption in combination with the desired environmental friendly solutions for power production development. Photovoltaic (PV) energy, which has gained a central place in the governments’ energy policies, is a promising solution to address the aforementioned concerns. Despite the advanced tools for PV power generation, several factors can significantly affect the PV system performance by decreasing its efficiency. Indeed, PV systems are frequently exposed to different sources of faults and failures that affect the power generated by the PV modules. These faults can considerably reduce its production efficiency and the lifespan of PV arrays. This is mainly due to external interferences or faults resulting from dust accumulation on PV modules, aging of PV modules, shading, MPPT error, and inverters faults. PV plants, especially its DC side might be subject to defects and anomalies causing a drop of the overall system performance or even their total unavailability. Therefore, a real time early fault detection and diagnosis procedure is crucial, not only for lowering maintenance cost and time, but also to avoid any energy losses, damage to equipment and safety hazards.