Data-driven Optimal Energy Management Strategy For Plug-in Hybrid Electric Bus

Plug-in hybrid electric vehicles(PHEVs)combine the advantages of both conventional hybrid and pure electric vehicles,so it is a hotspot for research and application with remarkable energy saving and emissions reduction effect.Energy management strategy is one of the key technologies for PHEVs,which has a significant effect on its fuel economy.This paper researches on data-driven optimal energy management,regarding four-mode hybrid power bus as research object.Firstly,energy management simulation model is built for four-mode hybrid power bus based on its structure and operating principle,which contains vehicle longitudinal dynamic model,power system model and driver model.The simulation model is verified by rule-based strategy.Secondly,driving condition data collection method for city bus is introduced,and the data is handled by SD card data analysis software and GPS coordinate transformation software to acquire GPS and CAN information.Meanwhile,energy management mathematical model for four-mode hybrid power system is established and analyzed according to optimization theory.Thirdly,the new characteristic parameter for representing driving condition segments based on pontryagin’s minimum principle is proposed.It is closely related to vehicle energy consumption and optimal control,comparing with statistic and PCA parameters which are calculated from vehicle speed trajectory.Then,data-driven equivalence factor estimation method is proposed based on the PMP parameter,and the results show that the average error account for 12.3% of the whole range,and the method can adapt to different driving conditions.Fourthly,data-driven optimal energy management strategy is established combining with driving condition database build and update method.Backward simulation,MIL test and HIL test are conducted to verify the strategy.The results show that energy consumption of the strategy is close to PMP optimum control and significantly better than rule-based control strategy about 6.3%,and the strategy also have good effect for different driving conditions.