Research On Modeling And State Estimation Of Lithium-ion Power Battery For Electric Vehicles

It is generally known that developing electric vehicles rapidly to establish green transportation system is not only the main way to solve the energy and environmental crisis but also the important direction of automobile industry transformation and development in 21 Century.With the strong support of national policy and the high concern of scientific research departments and the relevant enterprises,electric vehicles are meeting an unprecedented major development opportunity.The safe operation,mileage and energy management of electric vehicles are closely dependent on the rational use of power batteries.Because of the performance,capacity,service life and other obvious advantages,lithium-ion power battery has become the most widely used electric vehicle power battery.However,there are still many unsolved key problems in the efficient utilization and excellent cycle life of power batteries.As a result,Battery management system becomes so significant that it gained much attention.It becomes not only a hot issue in relative field but also a difficulty of technological development.These issues have become the main bottlenecks restricting the industrialization and practicability of electric vehicles,needing to be solved by studying new theories and methods.In this paper,the second-order RC model of lithium-ion battery and the fractional-order battery model based on fractional theory are established for the accurate modeling and state estimation of lithium-ion battery for electric vehicle.The precise modeling of the battery is realized.Based on the established model,the fractional Kalman filter method,the adaptive square root unscented Kalman filter and the model adaptive method are proposed to analyze the state of charge(SOC)and the state of health(SOH).The main works of this paper include the following aspects.1.Characteristics of lithium-ion power batteryThis part focuses on the construction of the power battery test platform,the establishment of the test database and the analysis of the characteristics of the power battery.Based on the AVL battery test platform produced by AVL in Austria,the lithium-ion battery of electric vehicle was tested,the test data was collected and the lithium-ion battery test database was established.The voltage characteristics,capacity characteristics,temperature characteristics and internal resistance characteristics of the lithium-ion battery are analyzed,which lay the foundation for the accurate modeling and state estimation of the lithium-ion battery.2.Research on modeling and parameter identification of lithium-ion power batteryAiming at the problem of modeling and parameter identification of lithium-ion battery,a lithium-ion battery model based on fractional theory is proposed and established.The computational complexity caused by too many RC modules and the problem of precise modeling of lithium-ion battery are solved.This paper studies the problem of fractional-order and parameter identification of battery model by genetic algorithm and provides a method to identify the order of the model.Finally,the accuracy and validity of the model are verified by experiments.It solves the trade-off between the complexity and precision of the dynamic lithium-ion model and fully reveals the fractional-order nature of the battery model.It provides an accurate battery model for battery state estimation3.Lithium-ion power battery SOC estimation based on fractional-order Kalman filter methodFirst,this paper deduces the state space expression of a discrete fractional-order system based on the Grunwald-Letnikov(G-L)fractional-order differential definition.Based on two lemmars,we derive the fractional-order discrete Kalman filter iterative formula.And the fractional-order Kalman filter method is used to estimate the state of the battery.The performance of the fractional-order Kalman filter method is verified by different working conditions.Compared with the method of traditional extended Kalman filter,the results show that the fractional-order Kalman filter method is faster and more accurate than the former.4.Lithium-ion power battery SOC estimation based on adaptive square root unscented Kalman filter methodAiming at the problem of accurate and reliable SOC estimation in the complex and variable environment of lithium-ion power battery,two methods of adaptive strong tracking unscented Kalman filter and adaptive square root unscented Kalman filter are proposed.In the traditional SOC estimation method,there are many disadvantages such as large estimation error,instability with noise.Therefore,this paper proposes two methods of adaptive strong tracking unscented Kalman filter and adaptive square root unscented Kalman filter to estimate SOC.The former introduces the fading factor in the state covariance matrix to adjust the error covariance in real time to weaken the influence of the battery equivalent circuit model misalignment on the SOC estimation.The latter can directly calculate the process noise variance of the lithium-ion battery system matrix and measurement of noise variance matrix,while ensuring that the state variance matrix and nonnegative noise variance matrix are non-negative and symmetry,thereby achieving noise adaptation and improving the accuracy,stability and adaptability of the algorithm.Finally,the validity of the two algorithms is verified by the experiment,and the accurate estimation of the SOC is realized by the lithium-ion battery in the complex and changeable environment and different working conditions.5.Lithium-ion power battery SOH estimation based on adaptive unscented Kalman filter methodAiming at the problem that the health state estimation of lithium-ion power battery,the immeasurable property of noise and the time-varying property of parameters,the SOH estimation method of lithium-ion battery based on model parameter adaptive is proposed.Firstly,the state space description of the battery system is established based on the second-order RC equivalent circuit model.An adaptive tracking-free Kalman filter method is proposed to identify the parameters of the model.Then,the mathematical relationship between the battery internal resistance and SOH is established.Noise adaptation was achieved by noise adaptive matching technology,and then the proposed algorithm is used to estimate SOH.Finally the effectiveness of the algorithm is verified through the test.In this paper,the quantum-cell modeling method based on fractional-order theory is proposed to solve the trade-off between the complexity and precision of the dynamic lithium-ion battery model,and the fractional-order essence of the battery model is fully revealed.The fractional-order Kalman filter algorithm and adaptive unscented Kalman filter method are studied to solve the problem of SOC and SOH estimation of power battery.It provides an effective and feasible method for battery state estimation for electric vehicle battery management system.In short,this paper revolves around the battery modeling,state estimation and other issues to carry out a series of studies.The results not only play a role in promoting for high-performance battery management system development,but also accelerate the industrialization of electric vehicles and practical process.