Aging State Estimation Of Lithium Ion Battery Based On Electrochemical Thermal Coupling Model

In recent years,lithium-ion batteries have been applied more and more widely in this field depending on their own advantages.However,battery performance deteriorates over time,and to ensure safe driving and avoid potential failures,estimate the state of health of the battery correctly become necessary.Researchers of battery materials often study the battery by disassembling the battery,so it cannot be directly applied in engineering;The data-driven battery health state estimation method needs a lot of data and does not involve the internal mechanism of the battery,so it cannot accurately describe the internal state of the battery.In this paper,an electrochemicalthermal coupling model is established,which contains many parameters that can reflect the internal mechanism of the cell.The method of obtaining model parameters was further studied,and the variation rule of electrochemical model parameters with aging was excavated,and some key parameters were used to complete the quantitative estimation of battery health state.Firstly,based on the pseudo-two-dimensional model,a discrete electrochemical model,including solid-liquid diffusion,charge conservation,Ohmic law of liquid phase and double electric layer,was established after partial simplification.The heat generation and heat dissipation conditions of lithium-ion batteries are analyzed,and a concentrated mass heat model which could output the internal temperature variation of the batteries is established.And the electrochemical-thermal coupling model is established.Due to the different effects of different parameters on the external characteristics of the battery,the sensitivity analysis of many electrochemical parameters was carried out.In the time domain,sensitivity analysis of parameters are carried out by changing a single variable based on the discrete degree of terminal voltage.In the frequency domain,electrochemical impedance spectroscopy combined with DRT analysis method was used to determine the polarization process sensitive to battery aging,which was used to guide the optimization of the electrochemical model and subsequent estimation of the state of health of battery.The parameters in the electrochemical-thermal coupling model were divided into different groups.The parameter values under different cycles were obtained by disassembling the commercial full battery for direct measurement,making the button half battery with the same material system,excite-response method,and genetic algorithm parameter identification.The condition of terminal voltage feature extraction and parameter identification is the custom multi-stage mixed pulse conditions at 25℃,and the experimental conditions are the CLTC-P standard road conditions.The root mean square errors of the results under different cycles are less than 0.03 V,which achieves a good fitting effect.Finally,the health state of the battery is estimated based on the parameters of the electrochemical model.The trend of electrochemical parameters changing with the number of cycles is analyzed qualitatively,and then the battery health status values under different number of cycles are calculated by the measured discharge capacity of the battery.Some parameters with high sensitivity and obvious variation with aging are selected from the electrochemical parameters to carry out correlation analysis with the battery health state.BP neural network is used to complete the quantitative estimation of the battery health state,and the estimation error is less than 1.52%...