Abstract:

Electric Vehicles (EVs) sales have grown at an exponential rate all over the world. However, the industry still faces many challenges like high cost of electric vehicles, range anxiety and lack of access to efficient charging infrastructure. With recent advancement in battery technologies and various subsidies provided by governments to incentivize electric vehicle adoption, lack of charging infrastructure remains the main issue. Moreover, an efficient network of charging stations will also reduce issues with range anxiety and reduce costs of vehicles as batteries constitute the major costs of an EV. The current charging infrastructure has many challenges like high upfront investment costs, low utilization rate of the chargers and high energy costs. Moreover, the charging stations need good management to lower peak demand and improve operation of the grid.  The objective of this thesis is to develop an intelligent control algorithm for mobile electric vehicle charging robots using reinforcement learning (RL) algorithms. The algorithm uses Deep Deterministic Policy Gradient (DDPG) based RL framework, which is an adaptation of Deep Q-learning on continuous action domain and uses an actor-critic and model-free algorithm on the deterministic policy gradient to operate over continuous action spaces.