Smart Control and Feasibility Analysis of Shared Electric Vehicle Charging Robots

Abstract

Electric Vehicles sales have grown at an exponential rate all over the world. However, this industry still faces many challenges with lack of charging infrastructure being the main problem. This study analyzes the feasibility of mobile electric vehicle charging robots being researched by industry and academia alike and proposes an intelligent control algorithm using deep reinforcement learning algorithms. The algorithm uses Deep Deterministic Policy Gradient based framework and uses an actor-critic and model-free algorithm on the deterministic policy gradient to operate over continuous action spaces. The charging solution is compared with existing conventional solutions using simulations. The results obtained from simulations show that a mobile autonomous charging station can provide many benefits. Apart from having a low upfront investment cost as compared to static chargers, a smart mobile charger also offers greater flexibility. The algorithm also performs better as compared to conventional algorithms like least laxity factor and can easily be adapted to recent trends like shared mobility and autonomous mobility to provide a better user experience.