A novel optimal power management strategy for plug-in hybrid electric vehicle with improved adaptability to traffic conditions

Adaptability to various driving conditions (TCs) is one of the essential indicators to assess the optimality of power management strategies (PMSs) of plug-in hybrid electric vehicles (PHEVs). In this study, a novel optimal PMS with the improved adaptability to TCs is proposed for PHEVs to achieve the energy-efficient control in momentary scenarios by virtue of advanced internet of vehicles (IoVs), thus contributing to remarkable promotion in fuel economy of PHEV. Firstly, the optimal control rules in the novel PMS, corresponding to diverse driving conditions, are optimized offline by the chaotic particle swarm optimization with sequential quadratic programming (CPSO-SQP), which can effectively endow the global optimization knowledge into the rule inspired method. Then, an online TC identification (TCI) method is designed by cooperatively exploiting multi-dimensional Gaussian distribution (MGD) and random forest (RF), where the MGD based analysis on the macrocosmic state of traffic contributes to valuable inputs for the RF based TC classification, and additionally the super regression ability of RF further improves the identification accuracy. Finally, the numerical simulation validations showcase that the novel optimal PMS can reasonably and instantly manage the power flow within power sources of PHEV under different TCs, manifesting its anticipated preferable controlling performance.