A review of curve fitting error criteria for solar cell I-V characteristics

Abstract

Various methods for recovering solar cell lumped circuit model parameters from experimental characteristics are briefly reviewed. The advantages of extracting parameters from illuminated characteristics are highlighted. These include the availability of accurate analytical expressions developed recently. A commonly used method of parameter recovery by curve fitting minimises σ which is defined as the r.m.s. of the relative current errors between the experimental and theoretical characteristics. This method is demonstrated to be unreliable when used with characteristics which have been collected by linear analogue to digital systems, or which have certain data point distributions. A more reliable minimisation criterion ε{lunate} is proposed. ε{lunate} is based on the area difference between the experimental and theoretical characteristics. Computation experiments show that the use of ε{lunate} results in much more accurate parameter recovery for both dark and illuminated characteristics, and that its accuracy is almost independent of data point distribution. ε{lunate} also provides a good basis for comparing the quality of fit of theoretical models to experimental characteristics. © 1986.