Comparison of Angstrom-Prescott, Multiple Regression and Artificial Neural Network Models for the Estimation of Global Solar Radiation in Warri, Nigeria

Abstract

In this paper, the application of artificial neural network, Angstrom-Prescott and multiple regressions models to study the estimation of global solar radiation in Warri, Nigeria for a time period of seventeen years were carried out. Our study based on Multi-Layer Perceptron (MLP) of artificial neural network was trained and tested using seventeen years (1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007) meteorological data. The error results and statistical analysis shows that MLP network has the minimum forecasting error and can be considered as a better model to estimate global solar radiation in Warri compare to the estimation from multiple regressions and Angstrom-Prescott models.

A Study of Conflict Issues Affecting Ugandan Housemaids who Migrate to Saudi Arabia and How to Mitigate them

March 10, 2026

This study aims to comprehensively analyse the complex interplay between