Modeling the Solar Radiation Parameter over Abuja using Neural Networks

Abstract

Many computer simulation models which predict growth, development and yield of agronomic and horticultural crops require daily weather data as input. One of these inputs is daily total solar radiation, which in many cases is not available owing to the high cost and complexity of the instrumentation needed to collect the data. In this work, a neural network model of the solar radiation over Abuja, Nigeria is developed. The model is useful for predicting the solar radiation intensity over Abuja, and this prediction is very important for many solar radiation applications, including gathering and validating information on the potential of Abuja for location of a solar energy generating station. The model used was developed using the Levenberg-Marquardt back-propagation algorithm with data from the Campbell automatic weather station situated at the University of Abuja, Nigeria. Results show that the predictions from the model developed generally agree with the observations. Midday solar radiation values typically exceed 600 W/m2 during equinoxes and sometimes drop below 500 W/m2 during solstices. Two indices (the Solar Radiation Diurnal Index (SRDI) and the Solar Radiation Annual Index (SRAI)) were also introduced in this work to respectively characterize the amounts of solar radiations received in each day and in each year for a given location. Relevant information is also provided to guide stakeholders for the location of a solar energy generating station in the region.
Keywords: Neural network, solar radiation, Abuja, prediction, solar radiation diurnal index, solar radiation annual index

Cite this Article
Iloanusi Nkiru Lilian, Okoh Daniel, Obi Joseph. Modeling the Solar Radiation Parameter over Abuja using Neural Networks. Research & Reviews: Journal of Space Science & Technology. 2017; 6(3): 40–48p.