Adaptive optimal control based on driving style recognition for plug-in hybrid electric vehicle

Abstract Vehicle energy economy is affected by different driving styles of individual drivers. To improve energy economy of plug-in hybrid electric vehicles (PHEVs), it is of great importance to develop the driving style adaptive optimal control strategy. In fact, driving styles are often influenced and restricted by different driving cycles. Therefore, to recognize driving style more accurately, this paper decouples driving styles from driving cycles. Based on classification and identification of driving cycles, the accelerator pedal opening and its change rate in different driving cycles are analyzed and the fuzzy-logic recognizer is built to identify driving styles. Afterwards, the driving style adaptive optimal control strategy is realized by combining the recognized driving style with the equivalent consumption minimization strategy (ECMS) and adopting a hybrid particle swarm optimization-genetic algorithm (PSO-GA) to optimize the relationship between the driving style and the equivalence factor (EF). The effectiveness of proposed driving style adaptive control strategy is validated by real vehicle test, which indicates that, compared with the original ECMS, the proposed driving style recognition based adaptive optimal control strategy improves the energy economy by 3.69% in the New European Driving Cycle (NEDC). This adaptive optimal strategy provides guidance for incorporating driving style into PHEV energy management strategy to improve fuel economy.