| The inevitable aging and performance degradation during the operation of lithium-ion batteries seriously threaten the long lifetime and high reliability requirements of electric vehicles(EVs)for power battery systems.Battery aging and life issues have become one of the bottlenecks restricting the large-scale development of EVs.Illuminating the aging mechanisms of batteries under various operating conditions,studying on the diagnosis method of battery aging and health states,as well as the evaluation method of battery lifetime are necessary for the durability management of power battery systems.The factors affecting battery aging are numerous and coupled with each other,and the aging states inside the battery are also difficult to directly measure,which will bring great challenges to the study of battery aging and life issues.This paper mainly takes the lithium ion batteries with Li(NiMnCo)O2 cathode(NMC)as the research object,and conducts cycling aging tests on the batteries under various conditions in the laboratory.Based on the accumulated large amount of battery degradation data,aiming at solving the durability management problem caused by battery aging,this paper will conduct in-depth research on the aging mode analysis,lifetime rapid assessment,state-of-health(SOH)diagnosis and aging mechanism modeling of the power lithium-ion batteries used in EVs.The main research contents are as follows:(1)Based on the incremental capacity analysis method and the phase transformation characteristics of positive electrode(PE)and negative electrode(NE)active materials,the corresponding relationships between the various partial regions enveloped under the incremental capacity curve(IC curve)of NMC battery and the different phase transformation processes of PE and NE are established.Then,the characteristic parameters that can reflect battery aging modes are extracted from the IC curve to establish the multi-indicators system of SOH for NMC battery,which will enrich the connotation of SOH characterization and lay the theoretical foundation for battery aging modes recognition.By analyzing the evolution of SOH characterization parameters with battery aging,the action mechanism of charging current and charging cut-off voltage on battery aging is clarified.That is,below the critical value,charging current mainly affects the loss of electrode active material(LAM)and charging cut-off voltage mainly affects the loss of lithium inventory(LLI).The reason for the different degradation behaviors for batteries cycled under 20%and 100%SOC depth is identified.That is,the batteries cycled under 20%SOC depth mainly suffer from LLI,while the LLI and LAMPE are comparable during the aging process of batteries cycled under 100%SOC depth.(2)Aiming at the issue of battery lifetime rapid evaluation,this paper identifies the equivalent relationship of degradation between the batteries cycled under 0-100%full SOC range and the batteries cycled under the five partitioned small SOC ranges with 20%depth.It is found that when suffering identical cycle times,the sum of the decrements of SOH characterization parameter related to LLI under the five partitioned SOC ranges is equal to that of cycled under 0-100%SOC range.The correlations between the SOH characterization parameters and battery capacity are analyzed.Based on the strong linear correlation between the parameters related to LLI and battery capacity,a multivariate linear regression model of capacity is established.The adaptability of the model to various cycling aging conditions such as different SOC ranges and different current rates is studied.In the last,based on the additivity of the decrements of SOH characterization parameter related to LLI under the partitioned SOC ranges and the strong adaptability of capacity regression model to different current rates,a novel battery lifetime evaluation method combining the cycling aging test under partitioned small SOC ranges with the accelerated aging test at high current rate is proposed,which can significantly shorten the time-consuming of battery lifetime test.(3)Battery cycle life tests are performed under various cycle conditions until capacity plummeting.To analyze the plummeting mechanism,an improved non-destructive diagnostic algorithm is proposed to quantify battery aging modes.Firstly,establishing the mathematical model of PE and NE equilibrium potential,and then based on this,the model describing the evolution of the matching relationship between PE and NE OCV-SOC curves as well as the mapping relationship of PE and NE curves to the SOC scale of full battery under various aging modes is established.Using this model,the OCV-SOC curves of fresh and aged batteries are constructed on the SOC scale of fresh battery.Battery aging modes can be quantitatively identified by minimizing the error between the constructed curves and the measured curves.By analyzing the evolution trend of battery aging modes before and after the plummeting,two plummeting mechanisms are concluded.Based on the multi-indicators system of battery SOH and the commonality of battery degradation under the two plummeting mechanisms,a characterization method of capacity plummeting with strong adaptability is proposed.By extracting the plummeting characteristic index from the cycling data,the online recognition of plummeting battery is realized.(4)Based on the P2D electrochemical model of lithium-ion batteries,this paper will establish the models describing battery dominant aging mechanisms that induce LLI and LAM,including SEI film thickening,electrode active material particles and SEI film fracture,lithium plating at NE.Integrating the established aging mechanism models with the original P2D electrochemical model,lithium-ion battery physical-chemical model taking into account the aging effect is obtained,which will lay the theoretical foundation for battery durability management.In order to solve the parameter identification problem caused by the fact that the model contains too many unknown parameters and they are coupled to each other,the parameters to be identified are classified according to the response characteristics of corresponding electrochemical process under specific excitation.And then,the identification scenarios of each type of parameters are designed separately to realize parameter decoupling identification.(5)The extended Kalman filter algorithm is used to online identify the evolution curve of battery OCV with the accumulated discharging capacity.Starting from fitting this curve,the effectiveness of the proposed battery aging mode quantitative diagnosis method applied in dynamic current discharge conditions is verified,indicating that this method can online diagnose battery aging modes,estimate its maximum available capacity and realize the regular updates of battery OCV-SOC curve along with aging. |
PDF Full Text Request