A General Model for EV Drivers’ Charging Behavior

Abstract

The increasing adoption of electric vehicles (EVs) due to technical advancements and environmental concerns requires wide deployment of public charging stations (CSs). In order to accelerate the EV penetration and predict the future CSs requirements and adopting proper policies for their deployment, studying the charging behavior of EV drivers is inevitable. This paper introduces a stochastic model that takes into the consideration the behavioral characteristics of EV drivers in particular in terms of their reaction to the EV battery charge level when deciding to charge or disconnect at a CS. The proposed model is applied in two case studies to describe the resultant collective behavior of EV drivers in a community using real field EV data obtained from a major North American campus network and part of London urban area. The model fits well to the datasets by tuning the model parameters. The sensitivity analysis of the model indicates that changes in the behavioral parameters affect the statistical characteristics of charging duration, vehicle connection time and EV demand profile, which has a substantial effect on congestion status in CSs. This proposed model is then applied in several scenarios to simulate the congestion status in public parking lots and predict the future charging points needed to guarantee the appropriate level of service quality. The results show that studying and controlling the EV drivers' behavior leads to a significant saving in CS capacity and results in consumer satisfaction, thus, profitability of the station owners. Index Terms-Electric Vehicle (EV), characteristic modeling, congestion control, electric vehicle charging. NOMENCLATURE CP Charging Point. CS Charging Station. M CS Monte Carlo Simulation. M DP Markov Decision Process. c i The state of the EV connection to the CP in Markov chain with i% level of charge. d i The state of the EV disconnection from the CP in Markov chain with i% level of charge. p iThe probability of EV driver's decision to disconnect with i% EV's level of charge, defined by logistic function. q iThe probability of EV driver's decision to connect while the EV is moving with i% EV's level of charge, defined by logistic function.tThe probability of EV parking in time t when the EV is disconnected from CP.

Foundation for adaptive charging solutions: Optimised use of electric vehicle charging capacity

“…From previous studies [1][2][3][4][5][6][7][8][9][10][11][12][13][14], only [4,14] mention nonideal charging characteristics, whereas the rest focus solely on EV usage-related behaviour. In [14], the realised charging currents are measured to determine the energy levels of the EVs more accurately.…”

An interleaved DC charging solar system for electric vehicle

“…Due to the limited and constrain power available in the battery for the EV, various studies have been conducted to prolong the battery life and maximizing the EV potential for long-run operation [3], [4]. The research on the EV system is not only limited to the components and control strategy in the EV system itself, the charging strategy is also another means towards the improvement of the EV for sustainability in the transportation sector [5], [6].…”

A probabilistic capacity planning methodology for plug-in electric vehicle charging lots with on-site energy storage systems

“…One important system parameter is customer arrival and charging demand statistics. In [30], using actual PEV charging data (level 2 chargers) from a major North American University Campus between 2010-2015, a Markovian model for representing the charging behaviour of PEV owners is presented. The results show that PEV owners not necessarily fully charge their batteries, hence, service duration are shorter than expected.…”