Recommendation and Prediction of Solar energy consumption for smart homes using machine learning algorithms

Abstract

Solar Energy is a principal source of renewable energy generation. Solar intensity is directly proportionate to solar power generation and it is highly reliant on the weather. A model is proposed that predicts the amounts of solar radiation produced using weather information implemented using various machine learning techniques such as Gradient boosting, SVM, etc. The results allow us to make effective energy consumption plans for smart homes with efficient utilization of solar energy which may provide several economic benefits. Additionally, accurate forecasts would make users more prepared to switch between conventional and renewable sources as required. A comparison study is performed with various machine learning models to determine the best method for building a prediction model. The groundwork for constructing models that could be dispatched to various regions is laid out that will incorporate that geographic location’s weather data, and output accurate solar intensity predictions for that area. Furthermore, a recommendation system is proposed for the consumption of thus predicted energy.