A Long Term PV Power Degradation Prediction due to Dust Fouling and Environmental Stresses

This paper presents a long term data driven power degradation prediction technique of photovoltaic (PV) module due to dust fouling along with different climatic parameters. Existing methods either analyze environmental effects or PV soiling separately. Combining the effects of temperature (static and cyclic), relative humidity (RH) and transmittance loss due to PV soiling, an integrated model has been developed. A nature based meta-heuristic optimization technique has been adopted to estimate the unknown model parameters. The model has been trained with a long term data (twelve years) taken from National Renewable Energy Laboratory (NREL) data bank and tested for subsequent two years. The measured data and predicted data are compared to establish the efficacy of the model, giving rise to encouraging results.