Learning and optimizing charging behavior at PEV charging stations: Randomized pricing experiments, and joint power and price optimization

In this paper, we introduce, implement, and assess a framework for jointly optimizing the pricing policy and the charging schedule of electric vehicles (EVs) by learning and shaping human behavior with pricing. The proposed methodology uses time-based pricing to incentivize user behavior at charging stations towards actions that achieve the station operator's objectives. The optimization framework incorporates endogenous human behaviors by explicitly accounting for the willingness to delay charging, as well as the plug-in duration of each session, as a function of the hourly prices. The approach also addresses the issue of overstay at EV charging stations by casting the problem as a trade-off between occupying resources and giving more flexibility to the station operator. We discuss the design and analysis of the behavioral experiments used to model the charging behavior of participants at the charging stations, and demonstrate the effectiveness of those learned behavioral models in an optimization framework aimed at minimizing the total cost of providing the charging service without sacrificing the user experience. Our simulations show that our framework increases total revenue, reduces utility cost, and increases net profit for the station operator, while maintaining a high level of service and consumer utility.

Automated summary

This paper introduces a framework for optimizing the pricing policy of electric vehicle charging by learning and shaping human behavior. The effectiveness of the framework is demonstrated through behavioral experiments and simulations.