Using mobility data and agent-based models to generate future E-mobility charging demand patterns

This paper presents an approach to generating electric vehicle (EV) charging demand patterns for residential areas, working locations, and public charging points for three different user groups and nine typical trip motives. We present a method to convert single-day trip diaries of a national travel survey to weekly vehicle mobility patterns. These weekly vehicle mobility patterns are used as input for a flexible agent-based model that describes the charging probability of an EV based on its State-of-Charge (SoC). The results are presented in a case study with 1,000 EV schedules and validated against the original dataset. The resulting approach is a feasible method to create EV demand patterns that can be used in a wide variety of research directions involving smart EV charging. The source code for the tool is publicly available.