Modelling driving and charging behaviours of electric vehicles using a data-driven approach combined with behavioural economics theory

Abstract

With the popularization and promotion of electric vehicles (EVs), their interactions with power grids and traffic networks have increasingly deepened. Accurate modelling of EV behaviour can faithfully depict the characteristics of EV driving and charging. However, most existing modelling researches fail to adopt real-world travel data and consider realistic perceptual decision-making psychology of owners. Thus, this paper proposes a novel behavioural modelling for EVs based on a data-driven approach combined with behavioural economics theory. To characterize the driving behaviour of owners using actual data, a systematic data mining and modelling approach is firstly proposed based on the open-source ‘Didi’ traffic travel data set, which obtains the traffic operation rules and the regenerative behaviour characteristics data. According to the subjective perceptual characteristics of social economic man, a Cumulative Prospect Theory-based modelling framework is further developed to quantify the uncertain and stochastic charging decision-making behaviour of EV users. Moreover, the user's preferences and attitudes are evaluated by calculating their cumulative prospect value when choosing charging stations. Finally, the most suitable charging station is recommended for EVs with charging requirements. Case studies are conducted within a practical zone in Nanjing, China. The results demonstrate that the traffic travel rules of vehicle owners have typical date types and functional area distribution characteristics. And the travel time and space of private and commercial vehicles are relatively regular, whereas the travel rules of public vehicles are random. Besides, this proposed methodology can not only effectively capture the irrational decision-making characteristics of EV users' charging behaviour, but also achieve promising performance in terms of reducing the charging waiting cost. The user's decision-making regarding charging behaviour exhibits a higher risk-seeking preference than a risk-aversion preference.