CIBECS: Consumer Input Based Electric Vehicle Charge Scheduling for a Residential Home

Electrical utility companies offer dynamic electricity pricing to limit peak demand of residential homes to provide charging for the fast-growing Electric Vehicle (EV) fleet. Charging EV at off-peak hours is economical for a user; however, scheduling brings the possibility of an undercharged EV at the time of use. The user has the best knowledge about his driving schedule, so including his input about target charge level and available charging time is an effective way to avoid such discomfort. To this end, this work proposes a Consumer Input Based Electric Vehicle Charge Scheduling (CIBECS) for a residential home. CIBECS takes consumer input, electricity price, and load forecasts to propose an adaptive scheduling technique. Moreover, we utilize an artificial neural network, particularly an LSTM network, to predict highly volatile residential loads. Experiments show our model's superior performance in minimizing electricity cost compared to existing approaches.