A GA-based online real-time optimized energy management strategy for plug-in hybrid electric vehicles

The performance of plug-in hybrid electric vehicles (PHEVs) depends on the energy management strategy (EMS). An optimal EMS can maximize energy utilization by coordinating the battery energy and fuel consumption. However, the real-time EMS remains a challenge due to the multiple dynamic parameters. To this end, this paper proposed a real-time EMS based on the adaptive regulation of multiple parameters including the driving cycle, driving distance and battery state of charge (SOC). First, the rule based EMS (RB-EMS) is designed, in which the operative threshold of the charge depleting mode (CD) and charge sustaining (CS) mode are initially determined through engineering experience considering the engine and battery SOC characteristics. Meanwhile, equivalent consumption minimization strategy (ECMS) is utilized to replace the traditional rules in the CD-CS region, which benefits to find the real-time optimal solution in a wider range and simultaneously addresses the fixed torque distribution between the engine and motor in the RB strategy. After then, to improve the adaptability of EMS in different driving cycles, the threshold in the RB strategy and equivalent factor (s(t)) in the ECMS strategy with different driving distances and initial SOC values are optimized with a global genetic algorithm (GA). Besides, a driving cycle recognition algorithm (DCR) based on fuzzy logic is online executed with GPS and onboard sensors, with which the optimal engine and motor torque distribution under the current conditions are adaptively determined according to the identified driving cycles and the above-mentioned offline results. Finally, the proposed adaptive hybrid energy management (HEM) approach was validated by the numerical simulation and hard-in-loop (HIL) experiments. Results show that the proposed HEM strategy can adaptively change the dynamic parameters to achieve the optimized CD stage in real-time; meanwhile, it can achieve a suboptimal result compared with DP which has reduced the fuel consumption by 14.82% for high power demand level. (c) 2021 Elsevier Ltd. All rights reserved.