BSO feature selection based machine learning solar radiation prediction

Abstract

Abstract Benefits of solar power production is constantly increasing to the electrical power grid. Renewable energy sources are becoming alternatives for energy resource around the world. In order to reduce environmental pollution and CO2 emissions, an ideal solution is provided to overcome the energy crisis. Renewable energy forecasting improves the accuracy and significantly improved by developing more solar forecasting models using numerical weather predictions. The solar radiation value reaching the system is very important in determining the energy production potential of the solar energy system. In this, we discuss the development of the project with machine learning combined with multiple metrological models to improve the accuracy of solar radiation forecasting. To implement combination of two models, Bird Swarm Optimization algorithm for select features and for classification Convolutional Neural Network is used. CNN is a system prediction which are including numerous atmospheric based on satellite images or several other weather prediction products.