Abstract: The driving condition adaptability method of instantaneous optimal energy management strategy was proposed to improve the fuel economy of plug-in hybrid electric vehicle (PHEV). The driving condition identification model was established based on back propagation (BP) neural network algorithm. The equivalent fuel factor sequence of standard driving condition was obtained by using dynamic programming algorithm under the constraints of battery power balance. According to the real-time identificated results of the model and the state of charge (SOC) of battery, the real-time application of the equivalent consumption minimization strategy (ECMS) was realized by using the interpolation method to solve the equivalent fuel factor at this time. The results shows that the proposed method can improve fuel economy and ensure battery power balance comparing with  the traditional equivalent consumption minimization strategy without considering the condition identification, and the fuel economy under 5 driving conditions are improved by 2.2%, 2.5%, 3.3%, 2.4% and 4.0%, respectively. 

Key words: plug-in hybrid electric vehicle (PHEV), equivalent consumption minimum strategy (ECMS), back propagation(BP) neural network algorithm, dynamic programming, driving condition identification