Baumgarte, FelixDombetzki, LucaKecht, ChristophWolf, LindaKeller, Robert2020-12-242020-12-242021-01-05978-0-9981331-4-0http://hdl.handle.net/10125/70719The widespread adoption of electric vehicles makes investments in charging parks both immediate and necessary to lower range anxiety and allow longer trips. However, many charging park operators struggle with sustainable and profitable operation due to high fees on peak loads and volatile availability of renewable energy. Smart charging strategies may enable such operation, but the computational complexity of most available algorithms increases significantly with the number of charging points. Thus, operators of larger charging parks need information systems that provide real-time decision support without immense cost for computation. This paper presents a model that uses recent methods from the field of Reinforcement Learning. Our model is trained on a charging park simulation with realworld data on highway traffic and day ahead energy prices. The results indicate that Reinforcement Learning is a feasible solution to improve the sustainable and profitable operation of large electric vehicle charging parks.10 pagesEnglishAttribution-NonCommercial-NoDerivatives 4.0 InternationalAnalytics and Decision Support for Green IS and Sustainability Applicationscharging parkdecision supportelectric vehiclereinforcement learningsmart charging10.24251/HICSS.2021.107