Electrochemical Reduced Order Modeling And SOC Estimation Of Lithium-ion Power Battery

With the increasingly serious problems of environmental pollution and energy shortage,the new energy vehicle has become the focus of the development of the automotive industry in 21 st Century.Power battery and battery management system are the key technologies that affect the development and popularization of new energy vehicles.Lithium-ion battery has become the most widely used power battery in the world due to its outstanding advantages.Under the complex dynamic driving conditions of the vehicle,the accurate estimation of the state of charge of the battery can greatly improve the service life and use efficiency of the battery.Compared with the equivalent circuit model,the electrochemical model can directly characterize the internal state of the battery and has higher estimation accuracy.However,the amount of calculation of the models is too large to realize the online estimation of the state of the power battery.In this paper,the electrochemical mechanism modeling and the reduction method of lithium-ion power battery are studied in depth.A lithium-ion battery SOC estimation algorithm,which promotes the application of the electrochemical model in the vehicle battery management system,based on electrochemical model is designed.The mainly work are list as follow:Based on the concentrated solution theory and porous electrode theory,the system structure of the pseudo-two-dimensional electrochemical model was analyzed.The electrochemical model of lithium-ion battery with highly nonlinear and strong coupling was preliminarily constructed.On this basis,a series of important assumptions are made,and three parameters parabolic method and Pad é approximation method are used to approximate the solid-phase and electrolyte-phase lithium-ion diffusion equations respectively.The simplified solution of the open circuit voltage,the electrolyte-phase potential and the over potential is completed,and a reduced order extended single particle electrochemical model is established.In order to further improve the accuracy of the model,a linear decreasing weight particle swarm optimization algorithm is applied to identify the key parameters of the extended single particle model.The particle swarm optimization algorithm is introduced in detail,including the principle of parameter identification,the improvement of inertia weight,the selection principle of parameters and the steps of parameter identification.Based on the 0.1C constant current discharge experimental data,the positive and negative electrode lithium-ion diffusion coefficient,the positive and negative reaction rate constant and the maximum particles lithium-ion concentration of positive and negative in the model are obtained.The extended single particle model of lithium-ion battery after parameter identification is built by Matlab/Simulink.The trend of the lithium-ion concentration of solid phase and electrolyte phase in the positive and negative electrode during the discharge process is analyzed,and the validity of the model reduction method is verified.The precision of the extended single particle model and the traditional single particle model is verified by the constant current discharge condition,the cycle pulse discharge condition and the federal urban driving schedule respectively.The experimental results show that the extended single particle model established in this paper has higher accuracy in all working conditions,and can better simulate the charge and discharge characteristics of the battery.The definition formula of the battery SOC in the electrochemical model is given.On the basis of the reduced order extended single particle model,the extended Kalman filter algorithm is used to estimate the lithium-ion concentration on the surface of the particle,and the SOC value of the battery is further calculated.The estimation accuracy of the algorithm is verified by FUDS dynamic test.Experimental results show that the SOC estimation accuracy of the proposed algorithm is high,and it has a good correction effect on the initial SOC error.