Parameter Identification And State Of Charge Estimation Of Lithium-ion Battery Based On Fractional Order Model

With the vigorous development of the new energy and electric vehicle industry in recent years,the lithium-ion battery plays an increasingly important role in the energy storage system（ESS）.In order to ensure the safe and reliable operation of the ESS,it is necessary to monitor the state of charge（SOC）of the battery.The fractional order model（FOM）can well reflect the external characteristics of lithium-ion batteries and improve the accuracy of SOC estimation.Therefore,the parameter identification and SOC estimation of single lithium-ion battery based on FOM are studied in this thesis.The advantages of the FOM compared to the second-order RC integer-order model are analyzed,and the accurate estimation of SOC is achieved.The specific contents of this thesis are as follows:In view of that the second-order RC equivalent circuit model cannot accurately fit the impedance characteristic of the actual lithium-ion battery,the impedance element with fractional calculus property is introduced and a variable order FOM is constructed.And then,the offline and online parameter identification algorithms applied to different models are introduced,and an adaptive forgetting factor recursive least squares（AFFRLS）algorithm is proposed.Through the adaptive forgetting factor formula,the rapidity and stability of the online identification algorithm is improved.Aiming at the difficult problem of identifying the FOM order online,a fractional order repeated prediction recursive least square（FORPRLS）algorithm is proposed.The order gradually converges to the optimal value in the recursive process by repeatedly verifying the terminal voltage error at the current moment.The dynamic stress test（DST）experiment is used to compare different parameter identification algorithms,and the superiority of the FOM and online identification algorithm in aspect of describing the battery dynamic characteristic is verified.Moreover,the effectiveness and strong robustness of the proposed algorithm are further verified by analyzing the prediction results and anti-noise effect of FORPRLS algorithm.The extended Kalman filter（EKF）is applied to the FOM online parameter identification algorithm,and the DST experiment is used to compare the SOC estimation results.The results show that the accuracy and convergence speed of online SOC estimation based on variable order FOM and FORPRLS-EKF algorithm are better than those based on fixed order FOM algorithm under different SOC initial values.The superiority of the constructed variable order FOM and FORPRLS-EKF algorithm is verified.Finally,the battery test system based on the bidirectional Buck-Boost converter is built.The SOC online estimation experiments of constant current charging,discharging and dynamic conditions are carried out.The feasibility and availability of FORPRLS-EKF algorithm are verified.