Balancing Charging Waiting Time and Traveling Time for Electric Vehicle Networks Using Reinforcement Learning

Abstract

Charging station selection policy for electric vehicle (EV) owners has received significant attentions as the number of EVs grows in recent years Two key factors concerned by EV owners are the distance and the queuing time associated with a charging station (CS) Applying reinforcement learning method while trying to balance the two factors this thesis presents two policies that help EV users to make a desired charging station selection Different environment settings are being considered and simulated using the celebrated traffic simulator SUMO The performance of the two proposed policies in comparison with two benchmark schemes policies are evaluated in terms of their ability to reduce the total amount of time EV owners need to spend on carrying out charging actions

Date of Award	2019
Original language	English
Supervisor	Kuang-Hao Liu (Supervisor)