Development of energy-optimal control strategies for a fully electric vehicle. Master of Science Thesis in the Master’s Programme Automotive Engineering

The development of new strategies in order to increase the range of fully electric vehicles by minimizing the energy consumption using on-board and off-board sources has been the subject of this work. To do so, an innovative energy management strategy has been developed for intersections regulated by Traffic Lights. EU founded research projects like OpEneR unify efforts of several partners to develop such new technologies. This work contributes to OpEneR project and deals with Traffic Lights. The objective is to develop energy optimal control strategies for the fully electric OpEneR prototypes. It is assumed that Vehicleto-Infrastructure (V2I) communication provides required data about Traffic Lights along considered road segment. Model Predictive Control is used as the framework methodology, which allows for specific powertrain representation and managing of the Traffic Light problem through constraints. In order to test the efficiency of developed control method, the controller is subjected to several representative simulation use-cases, including a real world road in Graz, Austria. The results have shown energy consumption reductions compared to a standard driver varying between 17 and 37 % depending on selected use-case. It is therefore concluded that the presented methodology constitutes an effective approach to the Traffic Light control problem management. Additionally, a Speed Advisory optimization-free algorithm dealing with Traffic Lights was also introduced. It is designed as a real-time implementable alternative to Model Predictive Control formulation. Furthermore, work performed with Human Machine Interface as well as a vehicle mock-up for demonstration purposes are also described. Finally, investigations regarding Explicit and Nonlinear Model Predictive Control are also presented and their strengths and drawbacks are highlighted. The Master Thesis has been written in close collaboration with AVL List GmbH and contributes to the European research project OpEneR.