Unified predictive fuel efficiency optimization using traffic light sequence information

Energy efficiency has become a major issue in trade, transportation and environment protection. While the next generation of zero emission propulsion systems still have difficulties in reaching similar travel distances as power-trains with combustion engines, it is already possible to increase fuel efficiency in regular vehicles by applying a more fuel efficient driving behavior. The rise of V2X technologies have opened up new possibilities for safety and energy efficiency applications. This publication proposes a model predictive approach that makes use of a power-train model and a sequence of traffic lights over a finite optimization horizon. The optimization problem is solved in a unified manner, i.e. power-train properties and traffic light phases are not considered separately but evaluated in a single cost function. A stagewise forward-backward Dynamic Programming approach involving cost re-utilization is used for optimization. In order to further decrease the search space, certain continuous entities are not explicitly regarded as a state component, but rather calculated during optimization.