| Abstrakt: |
Electrical power output variable from PV systems in outdoor conditions is substantially influenced by climatic input variables such as solar irradiance, wind, dust and temperature. In this work we proposed different types of modeling technique such as simple linear regression, multiple linear regression and non-linear regression for analyzing a PV systems. Then estimates of regression model’s parameters should be obtained accordingly, using some optimization methods like gradient method and Matlab function of regression model and machine learning, which relies on minimizing the sum of square of errors. Regression analysis describes the relationship between a dependent variable and several independent variables. An application on real data set is also provided from PV systems with three technologies polycrystalline, amorphous, and monocrystalline in polydisciplinary faculty of Ourzazate (FPO) - Morocco. The dependent variable consisted in electrical power, while the independent variables were the following: solar irradiance, ambient temperature and module temperature. A mathematical equation is used to estimate the electrical power. [ABSTRACT FROM AUTHOR] |
