Electric Vehicle Charging Recommendations Based on User Travel Demand

The electrification of transportation has become an inevitable trend for sustainable urban development. However, the rapid population of electric vehicles and the improvement of charging infrastructure are in a state of imbalance, urgently requiring solutions for the imperfect charging infrastructure, charging path decision-making, and charging time selection. Existing charging recommendations mostly rely on distances and charging prices, without considering the users’ travel demands. In this paper, aimed to maximize the utility of users’ travel, with charging as a constraint, we propose a user activity-based Markov decision Process (MDP). Besides, the availability of charging stations is a critical factor influencing the sustainable development of electric vehicles, we also consider the availability of charging station into this model and apply reinforcement learning algorithm to get the optimal charging recommendations. Finally, we provides a charging plan for electric vehicle users by extending user activity to a week.