A Novel Algorithm for Scheduling of Electric Vehicle Using Adaptive Load Forecasting with Vehicle-to-Grid Integration

This paper proposes a novel smart grid model to address the critical issues like load uncertainty, overload and clustered charging due to remarkable growth in the number of EVs.. A Multilayer perceptron (MLP) model is implemented to solve for the forecasting problem and is trained on historical load data along with meteorological variables like temperature, humidity, wind and rainfall. The TOU pricing scheme is used along with the vehicle to grid integration. User price preference-based system is deployed for optimal scheduling of electric vehicles, in which each user has different energy demand profile based on EV state of charge (SOC) and user preferences. Thus, dynamic pricing system promotes valley filling reducing the peak demand and shifting the EV load to off-peak hours minimizing the cost to the user. The intermittent power output of renewable energy resources is compensated by feeding electricity back to the grid using V2G method. This model improves grid performance by load shifting, optimal EV scheduling, valley filling. The model also helps in reducing electricity cost to end user.