Modeling EV charging choice considering risk attitudes and attribute non-attendance

In this paper, we developed and compared logit based models of Chinese electric vehicle (EV) drivers' charging choice behaviors in terms of whether or not to charge at a destination with a charging facility. First, we conducted a web-based stated preference survey to obtain EV drivers' charging choice data and EV-related risk attitudes. Using these data we developed EV drivers charging choice models based on hybrid choice modeling (HCM) framework, incorporating risk attitude and different decision strategies. The HCM binary logit model results indicate that the risk attitude variable is a significant predictor of EV drivers' charging choices. Comparing the two HCM latent class logit (LCL) models, the attribute non-attendance (ANA) LCL model gets a better goodness-of-fit over the fully compensatory LCL model. The ANA LCL model result shows that the EV drivers are divided into two classes: risk averse class focuses primarily on the amount of excess or buffer range they have available to complete their next trip; and risk seeking class balances price against their current state of charge. This paper provides more insights into charging behavior and can be used to estimate the charging demand.