Spectrally corrected direct normal irradiance based on artificial neural networks for high concentrator photovoltaic applications

Abstract

The electrical characterization of a HCPV (high concentrator photovoltaic) module or system is key issue for systems design and energy prediction. The electrical modelling of an HCPV module shows a significantly greater level of complexity than conventional PV (photovoltaic) technology due to the use of multi-junction solar cells and optical devices. An interesting approach for the modelling of an HCPV module is based on the premise that the electrical parameters of an HCPV module can be obtained from the spectrally corrected direct normal irradiance and the cell temperature. The advantage of this approach is that the spectral effects of an HCPV device are quantified by adjusting only the incident direct normal irradiance. The aim of this paper is to introduce a new method based on artificial neural networks to spectrally correct the direct normal irradiance for the electrical characterization of an HCPV module. The method takes into account the main atmospheric parameters that influence the performance of an HCPV module: air mass, aerosol optical depth and precipitable water. Results show that the proposed method accurately predicts the spectrally corrected direct normal irradiance with a RMSE (root mean square error) of 2.92% and a MBE (mean bias error) of 0%.