Optimal Operation of Residential EVs using DNN and Clustering based Energy Forecast

In this paper, we present a scheduling scheme for household Electric vehicles based on deep neural network based demand forecast. A novel clustering based Short Term Load Forecasting (STLF) using deep neural network (DNN) is presented in this paper to forecast the household and EV demand. The forecasting is performed on electricity demand profiles for 200 households from the Midwest region of the United States. Tensor-flow based deep learning platform was used to develop deep learning structure. The households are clustered according to demand profiles and the grouped consumers are used as the forecasting parameters. The scheduling model uses the forecasted household and EV demand values to develop a linear programming based optimization model to minimize the electricity cost for consumers. Household and cluster constraints are considered in the optimization model to limit the sudden surge in power demand during low-price time periods.