Machine Learning-Based Wind Power Prediction for Electric Vehicle Charging Hour Optimization

Renewable energy is the future of power generation and wind power is one of the important players in meeting the energy demand that is continuously increasing with the innovation of new technology like electric vehicles (EVs). In this paper, a machine learning-based approach has been considered to tackle the increased energy demand by predicting EVs' most convenient charging hour by renewable energy source. A wind turbine power generation dataset has been considered to predict power generation throughout the year. The study has been done in two parts, firstly we have applied ML algorithms and secondly, we have done time series analysis. Five ML algorithms have been used to compare the accuracy, correlation, and different types of errors. Among the five ML algorithms, XGB Regressor provided the best predictions for wind power generation to schedule EV charging with an accuracy of 98.17%. On the other hand, our time series study reveals that charging EVs between 1 AM to 6 AM is the most effective time for utilizing wind power and addressing energy demand.