Energy Management Strategy For EMT Vehicle Based On Driving Cycle Prediction

Nowadays,hybrid electric vehicles(HEV)have been widely developed because of the desire for the new energy vehicle from people.It has been generally accepted that the advantages of HEVs are low fuel consumption,low emissions,and without being limited by the range.Electro-Mechanicaltransmission system is an important part of HEVs.It is because of the participation of having more than one power source,the research of energy management strategy of HEVs has been the focus.To a great extent,the performance of HEVs depend on the energy management strategy.The energy management strategy studied in the paper is established on a dual mode electro-mechanical transmission system,it belongs to the PSHEV.The energy management strategy studied in this paper is very important to the dynamic performance and fuel economy of HEVs.The research object in this paper is electro-Mechanicaltransmission,the center of research is the energy management strategy.The driving cycle recognition based energy management strategy and the driving condition prediction based energy management strategy are deeply researched.The strategy based on driving cycle recognition and driving condition prediction are established.In the driving cycle recognition based energy management strategy,control system choose theoptimal control strategy according to the result of recognition.The chosen strategy will be on duty for the coming control time,when the control time comes to end,control system will be activated to determine the kind of control strategy again.In the driving condition prediction based energy management strategy,the ECU predicts driving conditions in a short future time according to the history information.Then optimize the control variables in predicting area.The result of optimization will be applied to the vehicle control.The prediction and the optimization keep working until to the end of the whole cycle.The advantages of HEVs which are low fuel consumption,low emissions will be enhanced.On the basic of the driving cycle recognition,the factors which will affect the recognition is studied.The kinds of characteristic parameters which can represent the driving cycle are deeply researched.The purpose of the research is to ensure both the recognition result is accurate and the computational burden must be small enough.On the basic of the neural network theory,the neural network is applied to predict the future driving information.The method to improve the accuracy of prediction is studied.In order to verify the effectiveness of the driving cycle recognition based energy management strategy and the driving condition prediction based energy management strategy,the EMT low-power function prototype bench test based on the 40 t level EMT vehicle is built.Experimental verification is accomplished,the experimental results show that the strategies established in this paper can realize controlling the electro-mechanical transmission system,there is a certain improvement when compared with some other control strategies on fuel economy.