---

This work focuses on the integration of turn costs in route planning for electric vehicles. Information regarding simulated vehicle runs is used to create consumption values for turn costs. An accordingly appropriate matching principle is described to link costs of brake or acceleration maneuvers extracted from this data to turns in graphs representing road networks. Next two types of graph extensions that can hold these values are presented, along with the corresponding methods required to generate them from graphs without turn costs. Further the alterations required by the Dijkstra routing algorithm to process these graph types are presented. The resulting optimal path computation algorithms, along with their different graph types, are then compared to each other in a series of experiments. Lastly a case study is used to show the various differences in routing with turn costs.

---

Alexandru Lesi, B.Sc.: Informatics student at the Karlsruhe Institute of Technology. Thesis written at the Institute of Theoretical Informatics, professorship Algorithmics I.

---