Using trip information for PHEV fuel consumption minimization

When driven past their all-electric range, plug-in hybrid vehicles (PHEVs) must use their engines. Numerous theoretical studies showed that the conventional control strategy, i.e. all-electric mode followed by a charge-sustaining mode, is not the most energy-efficient control strategy. Better strategies require knowledge of the trip ahead. In this paper, we present a method of predicting a trip for a given itinerary (vehicle speed, stop time, and grade) defined by using a geographical information system (GIS). For each segment of the itinerary, a vehicle speed profile is generated through a Markov process, defined by transition probabilities extracted from a large database of real-world trip records. Ten trip predictions are then generated from a single itinerary for evaluation of an optimal control strategy for a short-range power-split PHEV by using Autonomie, a powertrain modeling environment. The baseline controller uses rules and optimal operating point look-up tables when in charge-sustaining mode. The optimal controller uses the Pontryagin's Minimization Principle (PMP), the performance of which heavily depends on the choice of one scalar parameter, the equivalence factor. Finally, we demonstrate the fuel-saving potential of the PMP controller, using the aforementioned trip predictions.