Predictive Energy Management Control For Hybrid Electric Vehicle Based On Driving Style Recognition

The rapid development of the global automobile industry has brought about two major social problems: energy shortage and environmental pollution.Hybrid electric vehicles(HEVs)are the best choice to effectively solve these two problems.The advantages that HEV shows in terms of energy efficiency and range are largely due to the energy management strategies adopted.For non-autonomous vehicles,different driving styles of drivers lead to certain differences in fuel consumption.Therefore,the problem of designing an energy management control strategy for HEV considering driving style recognition was proposed in this paper.The research contents are as follows:Firstly,aiming at the problem that the driving style and the length of the prediction horizon affect the speed prediction accuracy and thus the fuel consumption in the free driving scenario,a predictive energy management control strategy based on driving style recognition was designed.In this strategy,principal component analysis and fuzzy Cmeans clustering method were used to cluster driving styles,and an appropriate prediction domain was selected by analyzing the prediction accuracy of the different prediction horizons.BP neural network was used to design the HEV speed prediction model.To achieve optimal power distribution,this strategy calculated driving power demand based on vehicle speed prediction and used a model predictive control method.Secondly,considering the actual driving scenarios of following the vehicle and crossing the traffic lights,aiming at the influence of driving styles on the optimal control of HEV energy management in different driving scenarios,a speed planning and energy management strategy based on driving style recognition was proposed in mixed driving scenarios.This strategy integrates network traffic information in a more comprehensive way.It used vehicle-to-vehicle and vehicle-to-installation to obtain the position of the preceding vehicle,distance,timing,and phase information of traffic lights,and distance information from the vehicle to the traffic light to classify mixed driving scenarios.In free driving scenarios,the following scenarios,and traffic light scenarios,the objective function considering driving safety was constructed to plan the host vehicle speed based on the speed prediction of the preceding vehicle.To allocate engine and motor power in this strategy’s energy management section,a sequential quadratic programming optimization algorithm was used.Finally,the effectiveness of the proposed energy management control strategy was verified on the Car Maker-MATLAB/Simulink co-simulation platform,and the advantages of the designed control strategy in reducing fuel consumption were verified by comparing it with other energy management strategies.