Online Parameter Identification And State Of Charge Estimation For Lithium-ion Battery

As a key technology of electric vehicles,power battery is widely researched all over the world,its precise estimation of State of Charge(SOC),however,is still a challenge in battery management system.SOC estimation impacts driving range of electric vehicles,battery energy management,security of battery and so on,therefore,the studies on SOC have important application value.The research work of the paper mainly include:First,research and design of the power battery module test.The test platform for battery is set up,and then a test scheme is designed.In order to obtain systematic and accurate battery test data,it is necessary to design the relevant battery test program according to the research purpose and establish a reliable test database by means of high-precision data acquisition equipment.The experimental data of power battery provided data support for the establishment of battery equivalent circuit model,SOC estimation model,battery capacity decline model and model accuracy verification.Second,establish a battery model and model parameters identification.By analyzing and comparing the commonly used battery models,an equivalent circuit model of the battery was determined and established.The method of identifying the off-line parameters and the method of online parameter identification were analyzed emphatically.Off-line parameter identification method is based on laboratory battery data,and utilizes non-linear fitting method to identify the model parameters,that is,at different SOC and different temperatures OCV,battery resistance,polarization capacitance,polarization resistance and other parameters value,and then use interpolation or fitting function to get parameter values at any SOC and temperature.Online parameter identification method is to use the real-time battery current and voltage as an input of forgetting factor recursive least-squares algorithm,the algorithm real-timely online estimates the parameters.Then,the battery test data were used to verify the reliability and accuracy of the model parameters obtained by the offline method and the online method,respectively.Third,SOC estimation for lithium-ion battery.Up to now,the majority of BMSs adopt open circuit voltage method,ampere-hour integral method,Extended Kalman Filtering method to estimate the SOC of battery,however,on account of the very strong nonlinearity of battery and the impact of noise,driving cycle and application environment,these methods have difficulty in precisely estimating SOC.For solving the problem,the paper adopts Adaptive Extended Kalman Filtering method(AEKF)that possesses noise covariance matching technique to estimate SOC.Innovation-based adaptive estimation is the core of AEKF method and make AEKF possess the strong capacity of noise suppression.Based on it,the paper ultimately designs AEKF method based on online parameters identification to estimate SOC.Forth,Lithium-ion capacity degradation model.The paper has given an outline of the method of Lithium-ion capacity degradation modeling,then based on accelerated life tests,the capacity degradation models under the stress of temperature,discharge rate and both of them are set up respectively.However,the use of battery is not constant,in order to still further accurately depict the process of battery capacity degradation,capacity degradation model under dynamic stresses is proposed.Then the model has been fused to SOC estimation,which solve the problem that imprecise battery maximum available capacity influence the SOC estimation performance.Fifth,hardware-in-the-loop(HIL)test.A hardware-in-the-loop test platform has been set up,and the designed SOC estimation algorithm would download in battery management system(BMS)then its reliability and availability would be verified.And then,battery bench test has been set up for verification of the SOC estimation method.