Research On Electrothermal Coupling Modeling And State Estimation Of Lithium-ion Batteries

The large-scale development of electric vehicles is an important strategic initiative for China to respond to the environmental energy crisis and promote its green development,which also becomes a powerful guarantee to achieve "carbon neutrality" and "emission peak".With the strong support of national policies,the electric vehicle industry has ushered in major development opportunities.As the main energy storage component of electric vehicle(EV),the performance of the power battery directly affects the EV dynamic and safety characteristics as well as its driving mileage.Presently,the lithium-ion battery is widely used in the new energy vehicle field owing to its superiorities in terms of high specific energy,no memory effect and long battery life.To ensure the battery’s safe and reliable operation,an intelligent and efficient battery management system(BMS)is indispensable.The BMS has become an attractive research hotspot since it can achieve functions including battery operation states estimation,battery equalization,fault diagnosis and charge/discharge management by monitoring the basic information such as voltage,current and temperature of the battery.The estimation for battery state of charge(SOC)and state of health(SOH)are the core technologies of the BMS,and which is also the key basis for battery charge/discharge control,life monitoring and vehicle energy management.However,accurate battery modeling and state estimation for SOC and SOH is still a formidable challenge due to the factors that it is strongly influenced by battery aging,environment temperature and complex electrochemical properties.Therefore,the main research and innovation are as follows:Aiming at the problem of the poor environmental adaptability of the battery model,an electric and thermal coupling model is firstly established for lithium-ion batteries,and the model parameters of the electrical and thermal models are identified respectively.Specifically,aiming at the problems of the long test cycle and expensive test equipment of traditional thermal model parameter identification methods,a fast identification method of battery thermal model parameters based on the discharge temperature rise curve is proposed.The proposed method only requires a set of constant-current discharge experiments with various discharge currents,which can obtain sufficient experimental data in a short time and significantly shorten the test cycle.Finally,the accuracy of the established model is verified with dynamic profiles at various environment temperatures ranging from 0℃ to 40℃,and the experimental results show that the root mean square errors of voltage and temperature predictions are less than 31 mV and 0.7℃respectively.Therefore,the electrical and thermal characteristics of the battery at different environment temperatures can be comprehensively and accurately described by the established model.Battery state of charge SOC indicates the remaining charge of the battery,and its accurate estimation is one of the core technologies of BMS.Therefore,according to the established electric and thermal coupling model,the strong tracking particle filtering based SOC estimation method is proposed to achieve accurate SOC estimation under different environment temperatures.Considering that battery operation is a process in which electrical and thermal characteristics are coupled with each other,and excessive charge/discharge power may cause battery overheating,which may lead to accelerated battery degradation or even cause safety accidents.As a result,the multiple parameters such as battery temperature,voltage and SOC are integrated to achieve the prediction for battery state of power(SOP)by introducing the temperature as one of the important constraints.Finally,the experimental verification for the effectiveness of proposed battery states estimation methods is conducted under dynamic conditions with various environmental temperatures.The results show that the proposed methods achieve an accurate estimation of the battery SOC over a wide temperature range,and also can effectively control the battery charge/discharge power under high temperature conditions,thus improving the safety of the battery system.Considering the lower state estimation accuracy of aging batteries,especially for the problem that the battery SOH estimation is greatly disturbed by temperature,a fusion estimation method for battery SOC and SOH is proposed in this paper with consideration of temperature and aging effects.Firstly,an offline and online combined parameter identification method is proposed to ensure model adaptability while significantly reducing the online computational complexity.Second,the adaptive unscented Kalman filter is adopted for the estimation of battery SOC and capacity,and the artificial neural network is employed to characterize the relationship of SOH with the estimated capacity and temperature,which effectively overcomes the influence of temperature on SOH estimation.Finally,the proposed fusion method is thoroughly verified with dynamic profiles at varied temperatures and aging statuses.The results show the superiority of the proposed method with the RMSE of SOC and SOH estimation less than 1.2%and 2%,respectively.Aiming at the poor consistency and low SOH estimation efficiency of retired batteries,a discharge current ratio based fast SOH estimation method is proposed.Firstly,the relationship between discharge current distribution and capacity in parallel connection is constructed based on the battery parallel analysis model,and the capacity of multiple single cells can be estimated simultaneously in one discharge cycle as well as the SOH of the battery can be determined.Then,the experimental platform is designed to validate the effectiveness of the proposed method,and the results verified its feasible,accurate and efficient performance.Finally,the experimental results for three types of batteries are conducted to analyze the effects of opencircuit voltage,battery parameters and load differences on the estimation accuracy.In summary,aiming to achieve safe and efficient management for lithium-ion batteries throughout their whole life cycle,this paper achieves breakthroughs in battery electro-thermal coupling modeling and parameter identification,multi-state estimation and fast estimation of retired battery health state,respectively.Importantly,a practical and efficient theoretical approach is proposed in this paper to provide theoretical and methodological support for the safe and efficient management of power batteries.