Smart Charging of Electrical Vehicles

In this talk, Prof. Zhou will present two papers on smart charging of electrical vehicles (EV). In the first paper, we consider the problem of a utility firm to price the charging service to incentivize customers to delay charging. Given the departure times of customers, we find the optimal charging policy for both the cases when customers arrive simultaneously and sequentially. We then construct algorithms to derive the optimal pricing scheme when the objective of the utility firm is to maximize either the social welfare or its own profits. We illustrate our results using models calibrated to data from PJM interconnect and examine the implication on carbon emissions. In the second paper, we consider the charging of EVs in a charging station (such as ChargePoint) with uncertainties in arrival and departure times and charging quantity to minimize the total expected costs. We formulate the problem as a steady-state stochastic program (SP) incorporating these uncertainties with an electricity tariff including demand charge. As the common approach suffers from the large-scale nature of this SP, we develop exponential cone program (ECP) approximations, which can be solved efficiently. For the case when the station has infinite capacity, we obtain the ECP approximation by leveraging the problem structure and provide a theoretical performance guarantee. For the case when the station has a finite capacity, we demonstrate the superior performance of our ECP approximation using numerical experiments calibrated to EV data from the U.K.

 

(The first paper is a joint work with Owen Wu and Safak Yucel; the second paper is a joint work with Chen Li and Long He.)