Fuzzy Model, Neural Network and Empirical Model for the Estimation of Global Solar Radiation for Port-Harcourt, Nigeria
Olumide Olufemi Akinnawo *
Department of Physical and Earth Sciences, Crawford University, Igbesa, Ogun State, Nigeria.
Oluwaseun Caleb Adebayo
Department of Physics, Federal University of Technology, P.M.B. 704, Akure, Ondo State, Nigeria.
Abel Giwa Usifo
Department of Physical and Earth Sciences, Crawford University, Igbesa, Ogun State, Nigeria.
Abiodun Kazeem Ogundele
Southwestern University of Nigeria, Okun-Owa, Sagamu-Benin Expressway, Ijebu-Ode, Ogun State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
The invaluable role of the estimation of global solar radiation in solar engineering systems provides very useful direction for various solar applications. This paper employs the Adaptive Neuro Fuzzy Inference System (ANFIS), Artificial Neural Network (ANN) and regressive technique for the prediction of global solar radiation(GSR) on horizontal surface using temperature swing and relative humidity as input parameters covering years 1981 to 2005. The performance of the models was tested using statistical indicators such as mean bias error (MBE), root mean square error (RMSE), and correlation coefficient (CC). The results with ANFIS and ANN method provide a relatively better prediction with ANFIS the more preferable.
Keywords: Angstrom model, fuzzy logic system, neural network, solar radiation, temperature.