Physical and Data-driven Hybrid Model for Outdoor Lifetime Prediction of PV Modules

When long-term photovoltaic performance degradation predictions are required after a short time, the existing physical and statistical methods often provide unrealistic degradation scenarios. Therefore, we present a new method to predict and forecast photovoltaic lifetime after small performance degradation. We combine a physical model with data-driven algorithms to achieve accurate predictions. To calibrate and validate our model, 18 photovoltaic modules that have been exposed for 35 years are used. Finally, the hybrid model has been benchmarked with the physical model to predict the lifetime of 3 experiment mono-crystalline modules installed in different climates. Improvements are visible with the new model.