Dynamical Diagnostic method for Solar Panels Health Maintenance

• Published in: 2022 International Symposium on Power Electronics, Electrical Drives, Automation and Motion (SPEEDAM) pp. 906 - 911
• Main Authors: Guesmia, Mohammed, Ghomri, Leila, Zerhouni, Noureddine
• Format: Conference Proceeding
• Language: English
• Published: IEEE 22-06-2022
• Subjects: Degradation; dynamic diagnostic; Heuristic algorithms; I(V) translation; Maintenance engineering; Power electronics; Prediction algorithms; Real-time systems; RUL; Schedules; solar panels
• DOI: 10.1109/SPEEDAM53979.2022.9842235

We propose in this article to explore a dynamic diagnostic approach that will allow to estimate the future degradation of a solar panel, and to determine the "Remaining Useful life time "RUL" according to the electrical energy demand. This algorithm uses the linear interpolation method "LIM" to translate the I(V) characteristic of the solar panel to the standard test condition "STC" in order to calculate the maximum power of the PV panel,and then applies a monthly random degradation to make a dynamic prediction using a linear regression that will adjust with each new monthly degradation to estimate the RUL. we tested the algorithm with a random degradation per month of a sum corresponding to 20% of the maximum power of the PV panel during 25 years which correspond to the reliability of the PV panel, we also added4 abrupt random degradation between 0.5% and 0.8% of the maximum power of the PV panel to try to simulate the occurrence of critical defects and see the impact on the prediction of the algorithm. The prediction of the algorithm for the first test converged 20 months before the maximum power of the PV panel became lower than the electrical energy demand, for the second test it converged 11 months in advance, but the algorithm made an error of 4 months due to the abrupt degradation that distorted the prediction. These results will allow us to contribute to a realistic Pronostic of PV panel lifetime and performance with real-time data to improve reliability and maintenance schedules.