A Control Strategy of Range Extended Electric Vehicles Based on Driving Condition Identification and Pontryagin ’ s Minimum Principle

Range extended electric vehicles (REEVs) provide potential to increase driving mileage and lower the fuel consumption compared with common hybrid electric vehicles (HEVs). To distribute the power between the auxiliary power unit (APU) and the energy buffer (normally is a battery) without sacrificing fuel economy, many optimal control strategies have been proposed. Some strategies, however, seldom take the driving condition into account which could influence the consequence significantly of the optimal control strategies. Thus, this paper firstly evaluates the statistical feature of typical driving cycle with novel method, then by applying the learning vector quantization (LVQ) network, the real-time driving condition can be determined. According to specific driving condition, the Pontryagin’s minimum principle (PMP) as a global optimal solution is used to distribute the power. Simulation study proves the proposed control strategies should be a possible solution with reasonable viability. Keywordsoptimal control; range-extended electric vehicle; learning vector quantization network; Pontryagin’s minimum principle; driving condition identifying