Online Fault Detection and Diagnosis Methodology for Photovoltaic Systems Using I–V Characteristic Analysis

This paper introduces an advanced online fault detection and diagnosis (FDD) methodology for photovoltaic (PV) systems based on I–V characteristic analysis. While traditional I–V analysis is effective in evaluating PV system performance, it is limited by its reliance on offline detection, which can disrupt normal power generation. The proposed methodology addresses this limitation by employing an innovative online approach that utilizes real-time data from the monitoring system. Despite certain constraints, this approach processes the captured data to generate corrected parameters, estimated values, and reference values through data preprocessing. This enables the rapid identification of various faults, including open-circuit faults (OCFs), short-circuit faults (SCFs), shading faults (SFs), and hybrid faults involving SCFs and SFs. Additionally, it allows for the assessment of shading degree, number of short-circuit modules, and degradation degree. The method overcomes the limitations of offline I–V analysis and offers advantages such as real-time operation, low cost, and rapid fault detection. To verify the effectiveness, accuracy, and generalization capability of the proposed method, a comprehensive validation framework was conducted, including theoretical derivation, MATLAB/Simulink simulations, real fault experiments, and topology extension tests. The simulation and experimental results demonstrate that, for irradiance levels of 200 W/m² and above, the proposed FDD framework can reliably and accurately identify all predefined fault categories. This approach represents a substantial improvement in PV system maintenance and reliability, enabling more efficient and timely fault detection.