Assessing Photovoltaic Performance using Local Linear Quantile Regression

P.E. McSharry (UK)


Solar energy, Photovoltaic, Performance, Quantile regres sion


The power generated by eleven different photovoltaic tech nologies and meteorological variables, such as irradiance and temperature, were recorded every half-hour at sites in the UK and Spain. These photovoltaic technologies included monocrystalline, multicrystalline and amorphous silicon, copper indium diselenide and cadmium telluride. Local linear quantile regression was used to determine the conversion efficiency of each technology as a function of the irradiance. This non-parametric technique also pro vides confidence intervals for the estimates. Monocrys talline silicon had the highest conversion efficiency, rang ing between 10% and 13%.

Important Links:

Go Back