Date of Publication :22nd February 2018

Abstract: In this paper, we attempt to compare the output prediction accuracy of two models – the ANN model and the Linear Regression model as applied to the case of the simulated performance of the PV module. The input variables for each model are Voc, Isc, Insolation level, Temperature while the output variable is Pmax. ANN models after trial show that the model with 4-10-1-1 configuration is suitable. The linear regression model is multiple regression models involving the same variables. The data base for this study is derived from various available manufacturers’ data, together with some experimental outputs from a few research reports. The results of this analysis appear to indicate that the ANN model is superior in prediction performance.

Reference :