Theoretic and experimental performance of a grid-connected photovoltaic system: Multiple prediction model of efficiency and annual energy generation

Abstract

In this article, the performance of a 3.36 kWp grid-connected photovoltaic system (GCPVS) under warm and subhumid weather conditions and the development of a predictive mathematical model is presented. Climate data of the 2021 year were used to evaluate energy generation, different types of performance, and efficiency. The average annual yield, corrected yield, array, and final yields were 6.45 h/day, 6.18 h/day, 5.16 h/day, and 4.97 h/day, respectively. The overall annual mean capacity factor and efficiency ratios were 20.73% and 77.22%, correspondingly. Experimental data were analyzed and correlated by multivariate linear regression (MLR) prediction and simulation to validate models. The MLR analysis showed that the efficiency is highly dependent on the temperature of the PV modules and that climatic parameters significantly affect the efficiency and output electric power. The prediction models for PV module efficiency, system efficiency, and direct current energy exhibit an uncertainty of ±1.04%, ±0.57%, and ±35.38 kWh, one-to-one. The monthly generation was compared with results obtained by Energy3D simulation-free software, showing an absolute error of ±2.33 kWh. This information can be used as a methodological tool for predicting efficiency and power generation in direct current.