Forecasting of Solar Radiation of an Area by Using Artificial Neural Network Technique through Matlab

Abstract

Over the years, solar radiation is the area of major concern. Most of the inventions done
in last few years, are in renewable energy generation as conventional energy sources are
not enough for years to come. There is a need of finding alternatives of nonrenewable
sources. This study is totally based on solar energy to give some contribution to the
development of renewable energy generation. In this paper, a detailed analysis has been
made for forecasting of solar radiation by using Artificial Neural Network. In this
Forecasting method the inputs are taken by using different sensors, which are connected by
Ardunio UNO board. The collected data then stored in a excel sheet by a software called
PLX-DAQ, which represents collected data into a graphical form for further studies. Now
this data can be used for developing an ANN model into MATLAB by using ANN toolbox. The
network that is developed in MATLAB can be trained and sufficient information can be
extracted from it. Compressions can be carried out with the original data and trained data
and the results now shows rate of error between both. This will also give the forecast
accuracy i.e. obviously improved under variable weather conditions.