Energy Management Strategy For Hybrid Electric Vehicle Based On Neural Network And PMP

In the context of the energy crisis and environmental pollution,the performance advantages of HEVs have made them the focus of current research.The energy management strategy is the key technology to a hybrid vehicle.It can increase the fuel economy and reduce emissions by coordinating the energy distribution between the two energy sources.The research object is a single-shaft parallel hybrid electric vehicle.To meet the requirements of vehicle dynamic performance,the purpose of this study is to reduce the fuel consumption in hybrid-electric vehicle operation.An energy management strategy is designed for optimizing energy allocation in hybrid vehicles in real time energy management.The specific work is as follows:?1?According to the characteristics of single-shaft parallel hybrid vehicle power-train system,a complete vehicle model of hybrid vehicle is built based on Matlab/Simulink software platform,providing a necessary simulation condition for developing an energy management strategy.?2?Rule-based energy management strategies and energy management strategies based on Pontryagin's minimum principle?PMP?.Firstly,to extend the battery service life,the SOC of the battery is used as the main reference and a rule-based energy management strategy is established.Secondly,the control variable P_batt and the state variable SOC were determined to develop the objective function of the hybrid power-train system,and an PMP-based energy management controller was successfully designed.?3?Energy management strategy for hybrid electric vehicles based on neural network.The clustering algorithm is used to classify and train the data sets obtained from the PMP-based energy management strategy,and a feed-forward neural network controller is built.The results show the designed energy management strategy is reasonable,which maintains better real-time control performance and also meets the fuel-economy requirements.