A real-time capable enhanced dynamic programming approach for predictive optimal cruise control in hybrid electric vehicles

Abstract

Hybrid electric vehicles (HEV) combine both a powerful engine and the ability to reduce fuel consumption through an electric machine by intelligent control of a complex drive train. The driving strategy controls the speed and the operating strategy controls the energy management of the HEV. Both are combined in one optimization. The result is a driver assistance system that is tuned with predictive information about the upcoming route. A novel algorithm based on dynamic programming (DP) is presented that allows real-time application by introducing heuristic model-based optimization space reductions. The main focus is on the design and dimensioning of a slim DP algorithm with the capability to find an optimal control in close real-time.