Stationary Spatial Charging Demand Distribution for Commercial Electric Vehicles in Urban Area

Xinwu Qian1, Jiawei Xue2, Stanislav Sobolevsky3, Chao Yang4, Satish Ukkusuri2

  • 1PUrdue
  • 2Purdue University
  • 3New York University
  • 4Tongji University


11:30 - 11:45 | Mon 28 October | Gallery Room 1 | MoC-T9.3

Session: Regular Session on Electric Vehicles and Mobility (I)

Category: Regular Session


The electric vehicle (EV) is an increasingly popular mobility solution today. It is especially appealing for urban commercial usage including serving commuters and delivering goods. Commercial EVs, defined as electric taxi fleets in this paper, contribute a lot to the energy consumption and emission. One critical component to facilitate the adoption of EVs is to provide sufficient charging infrastructures for road users to alleviate the range anxiety. In this study, we develop an analytical model to profile the stationary spatial distribution of charging demand for any given fleet of commercial EVs, which serves as the necessary input for planning the location of charging infrastructures. Our model considers the movement of EVs as a random walk, with the underlying Markov process driven by the spatial demand that the commercial EV fleet needs to serve. We then define the electricity consumption and charging models and develop the probability distribution that an EV may need to recharge its battery at any given location in the city, based on the stationary distribution of the random walk process. We calibrate the performance parameters of EVs from the trip record and present numerical experiments based on New York City taxi trip data and compare the solution from simulation with our analytical solution. The result shows that analytical solution is quite close to the simulation result. This paper provides insights to the city agencies and private companies to make charging infrastructure location planning.