Design Of Rapid State Evaluation System For Lithium-ion Battery Based On Impedance Spectroscopy

Lithium-ion battery has played an increasingly important role in all aspects of human production and life since it was born.For the state estimation of batteries in some special application scenarios,it is necessary to find a fast evaluation method which is different from the traditional state estimation method.The degradation of lithium-ion battery performance is reflected in its impedance,so this paper builds a time-domain measurement system based on the electrochemical impedance spectroscopy of lithiumion battery,and then carries out modeling analysis and parameter identification based on the impedance spectrum,and tests the simulation ability of the model under different working conditions,designs estimation algorithm of SOH(State of Health)and SOP(State of Power)of lithium-ion batteries.The specific contents of this paper are as follows:First of all,in order to realize the fast and accurate measurement of the impedance spectrum of lithium-ion batteries,a time-domain measurement system of the impedance spectrum based on the fast Fourier transform is designed,including the design of hardware circuit and software flow.And the measurement accuracy is verified by measuring the impedance spectrum of the Li Fe PO4 battery,and the measurement results of the equivalent circuit impedance spectrum and the electrochemical workstation are used to check the measurement system Accuracy;the same state of the battery to carry out multiple measurements to verify the consistency of the measurement results;verify the effectiveness of the measurement system for different types of lithium-ion battery impedance spectrum measurement.Secondly,the second-order and third-order fractional order impedance models are established for the impedance spectra of Li Fe PO4 18650 battery and lithium Li Co O2 UR14500 battery respectively;the relationship between the model impedance spectrum and the model parameters is established by mathematical analysis method,and the validity of the analytical method is verified;On this basis,a fusion identification algorithm based on mathematical analysis and Levenberg-Marquardt algorithm is proposed,The parameter identification process of Levenberg-Marquardt algorithm is designed,and the accuracy of the identification method is verified based on the actual battery impedance spectrum data,and the error levels of amplitude and phase angle are analyzed.Finally,the second-order and third-order fractional order models are established for the two kinds of batteries,and the frequency-domain/time-domain transformation of the models is carried out with the help of fractional order calculus tools.The simulation ability of terminal voltage and the simulation accuracy of capacity under the condition of constant current are verified respectively;The impedance spectrum,battery capacity and charging SOP of Li Co O2 battery at different aging stages were measured.The SOH estimation algorithm based on BP neural network and model parameters is designed,and the accuracy of SOH estimation is verified;the SOP simulation estimation algorithm based on dichotomy is designed to analyze the error level;finally,the upper computer interface of battery state assessment system is designed and debugged based on all the above work.In this paper the fast and accurate measurement of the battery impedance spectrum is realized,the fractional model modeling and parameter identification based on the impedance spectrum are solved,the simulation accuracy of the model is verified,the state estimation algorithm is developed,and the interface of the state evaluation system is designed.The work done in this paper has a certain engineering application value in the rapid assessment of battery state.