A Nudge-Based Approach for Day-Ahead Optimal Scheduling of Destination Charging Station With Flexible Regulation Capacity

This work introduces the nudge theory of behavioral economics in electric vehicle (EV) charging schedules. Nudges emphasize guiding user choice through noncoercive means that go beyond economics. First, we conduct the surveys of EV users to verify the effectiveness of nudges in guiding charging behavior and establish user response models for day-ahead simulation. Then, we build a charging agent that can realize the rapid optimization of a single EV to reduce the electricity cost and improve the flexible regulation capacity (FRC). This agent is modeled and trained based on twin delayed deep deterministic policy gradient (TD3) algorithm. Next, we perform Monte Carlo (MC) simulations with this agent to calculate the day-ahead variability range of power consumption and FRCs involving various EVs. Finally, we set the price difference of the charging options with nudges as the control variable and utilize discrete particle swarm optimization (DPSO) to obtain its optimal value with the objective of the revenue of the charging station operator (CSO). The case study is based on over 1500 questionnaires. The results highlight that the nudges can improve the percentage of users participating in scheduled charging, thereby helping CSO achieve more revenue and offer reliable FRCs to the grid.