Study On Parameter Optimization And Energy Management Of Hybrid Energy Storage System For Electric Vehicles

With the environmental pollution and energy crisis becoming more and more serious,electric vehicles have been developing rapidly.Lithium-ion battery is the most promising candidate as on-board power source for its high energy density and mature production technique.However,low power density,short cycle life and strictly limited charging current of lithium-ion battery would restrict the vehicle performance,and bring about a series of problems,for instance,battery pack are aging fast,the pack maintenance costs are prohibitive and braking energy recovery rate is unsatisfactory.Hybrid energy storage system(HESS)is the technology that combines battery with ultracapacitor to supply power for the vehicle together.Ultracapacitor is a kind of electrochemical equipment with ultrahigh power density,millions of time of cycle life and almost unlimited charging current,which could compensate for the defects of battery.The combination of battery and ultracapacitor enhances both advantages as well as avoids both disadvantages,the HESS exhibits high power density and high energy density simultaneously;furthermore,as the ultracapacitor assists battery as an energy buffer,the battery life is obviously prolonged.Focusing on the HESS of EVs,and for the purpose of maximizing the performance of the HESS,this paper researched HESS from four aspects: the components performance,the topology selection the HESS,parameter matching of the HESS and the energy management development,the detailed research is shown below.The performance tests of the battery and the ultracapacitor are carried out from four aspects: open circuit voltage characteristic,actual capacity,ohmic resistance and polarization phenomenon.The mentioned characteristics are represented by lumped parameter equivalent circuit,furthermore,the lumped parameters are fitted by the least square method.Based on the fitting data,different HESS topologies are analyzed and compared,and the topology for target vehicle is chosen.On the basis of testing and the chosen topology,a combined multi-objective parameter optimization and energy management optimization method for the HESS is proposed.The Controlled Elitist NSGA-? algorithm is introduced to solve this problem,whose optimization goals are HESS replacement costs and vehicle economy.The controlled elitism strategy allows individuals who perform worse in the targets to be reserved in a certain ratio.In this way,the prematurity of NSGA-? algorithm is overcome.NSGA-? optimizes both the HESS parameters and the control strategy by calling the Simulink model of the HESS,obtaining the Pareto sets for both targets.Furthermore,by comparing the HESS initial costs,the vehicle average daily costs and the battery pack lifetime range of the Pareto sets,optimal HESS parameters are finally decided.In order to reduce computation burden,as well as to ensure the optimality of the parameters,convex programming(CP)method is introduced to solve this combined optimization problem.Firstly,the battery model is linearized,then the optimization problem is translated into convex optimization problem by introducing new variables.The optimization objective is the weighted sum of HESS initial costs and the battery pack capacity losses,both HESS parameters and optimal power split between battery and ultracapacitor are obtained by CP method.The proposed CP method reduced the computation burden of the problem,the result is optimal.A detailed HESS model is built in Matlab/Simulink platform.The HESS energy management strategy(EMS)is designed based on adaptive model predictive control(AMPC),minimizing both battery current and its rate online.The EMS consists of control-oriented module construction,model states estimation based on time-varying kalman filter and the quadratic programming solver for the AMPC.Compared with standard MPC,AMPC exhibits better estimation accuracy,compared with SMPC,AMPC is superior in controlling nonlinear model.The proposed AMPC is also compared with rule-based controller(RBC)and dynamic programming,results show that AMPC reduces Ahthrouthput by averagely 10.6% when compared with RBC,showing superior control optimality and adaptation to the driving condition uncertainty.