Fuel Cell Hybrid Vehicle Energy Management Strategy Based On Vehicle Speed Prediction

As fuel supply and demand problems and environmental pollution problems become more and more serious,new green-powered electric vehicles have attracted widespread attention from scholars at home and abroad.The Fuel Cell Hybrid Electric Vehicle(FCHEV),which is powered by Proton Exchange Membrane Fuel Cell(PEMFC)and lithium-ion battery,has become one of the most promising new environmentally friendly vehicles because of its zero emission,high efficiency,good regeneration and wide range of fuel sources.Since the dynamic response of PEMFC is slow and energy cannot be recovered,it is necessary to design a reasonable energy management strategy(EMS)to coordinate the power output of two energy sources,to meet the requirements of vehicle dynamics while improving the economy of FCHEV,to achieve energy saving and to extend the fuel cell life,which is important for the performance improvement of FCHEV.It is of great significance to improve the performance of the whole vehicle.This paper takes the power system of Dongfeng X37 model as the research object and the following work is done in this paper:Firstly,the advantages and disadvantages of various fuel cell hybrid power system structure types are compared,and the topologies of PEMFC and lithium-ion battery are selected,and the vehicle longitudinal dynamics model of FCHEV is established with reference to the data of Dongfeng X37 model,and the motor model and lithium-ion battery model are established in turn,and the causes of voltage drop generated by different polarization phenomena of PEMFC are analyzed in detail.It can be seen that the PEMFC polarization phenomenon leads to the soft output voltage characteristics of PEMFC,and the efficiency increases and then decreases with the output power,and then the output voltage model and hydrogen consumption model are derived.The model of fuel cell hybrid vehicle is established to lay the foundation for the subsequent design of EMS to make the fuel cell work in the high efficiency range and improve the economy.Secondly,a radial basis function(RBF)neural network-based vehicle speed prediction model is developed to address the problem that the existing EMS cannot predictably distribute energy to FCHEVs and cannot adapt to all operating conditions,and can only be optimized offline with poor economy.The Sparrow Search Algorithm(SSA)is used to adjust the initial clustering center of the network,and the training speed of the neural network is improved by updating the center value,width value and weight of the RBF neural network model.Based on the standard driving cycle combined working conditions as training data,three types of working conditions,namely,typical urban flow condition,urban congestion condition and highway condition,are selected as test samples,and the validity of speed prediction with different prediction lengths is verified,and the relationship between the accuracy of the prediction model and the prediction length is analyzed with root mean square error as the evaluation criterion,which verifies the accuracy and reasonableness of the designed SSA-RBF neural network vehicle speed prediction model The accuracy and reasonableness of the designed SSA-RBF neural network speed prediction model are verified,and it can be effectively used for online EMS formulation.Finally,Pontryagin’s Minimum Principle(PMP)is applied to solve the problem of optimizing the global energy distribution of FCHEV by solving the Hamiltonian function to obtain the optimal PEMFC output power,distributing the power output between the lithium-ion battery and PEMFC by co-state.The inequality constraint is introduced to prevent the drastic change of PEMFC output power and improve the durability of PEMFC.Simulation tests of PMP under UDDS operating conditions show the important influence of covariance values on PMP performance.To obtain the optimal co-state variable to solve the online application problem of PMP for different working conditions,an adaptive PMP strategy is constructed,which is combined with the SSA-RBF vehicle speed prediction method to achieve real-time online adjustment of the co-state.By conducting a comparative study under three cases of known future working conditions,unknown future working conditions and predicted future working conditions,the effectiveness of EMS based on vehicle speed prediction is verified,and the stability of FCHEV operation is improved by ensuring the smooth change of lithium-ion state of charge(SOC).The adaptive PMP strategy is compared and analyzed with the rule-based EMS under different operating conditions to further verify the effectiveness and practical application potential of the proposed EMS based on vehicle speed prediction in improving the fuel efficiency of FCHEVs.The strategy can not only stabilize the trajectory of the battery charge and discharge state and improve the safety and stability of FCHEV operation,but also significantly reduce its hydrogen consumption and substantially improve its energy management efficiency.