Solar PV Data Analysis

Solar photovoltaic (PV) systems are increasingly recognized as crucial sustainable energy sources with diverse applications. Their implementation leverages rapid advancements in material engineering, communication systems, and computational intelligence tools. This paper focuses on mathematical methods for signal analysis, including multichannel signal processing, optimization methods, and feature evaluation, to monitor PV systems with panels situated in a specific coordinate system. In particular, it examines systems with east/west oriented photovoltaic panels, employing computational tools to analyze power signals, assess time and positioning data, evaluate symmetry coefficients, and estimate features for detecting potential panel failures. The proposed methodology incorporates digital filtering, numerical and statistical methods, and clustering of multichannel power energy features. Additionally, a general graphical user interface for data analysis is proposed. A detailed case study is presented, analyzing the distribution of selected features over 70-day segments of a PV system comprising 7 east-oriented and 7 west-oriented panels, with data recorded over 5 days at 15-minute intervals. The results reveal distinct and well-separated feature clusters for healthy PV panels. The conclusions underscore the effectiveness of general signal processing tools in enhancing the integration of PV systems into existing grid systems, thereby improving their efficiency and facilitating the interconnection of monitoring systems with communication technologies.