Research On Remaining Useful Life Prediction Of Lithium-ion Battery

Lithium ion battery is the main energy source of electronic equipment and system.When the lithium ion battery fails or breaks down,it will bring serious accidents,cause economic losses,and even threaten the safety of users.If the remaining useful life(RUL)under various working conditions can be accurately predicted in the cycle working process of lithium-ion battery,the battery can be replaced before the battery failure or failure,so as to reduce the probability of accidents caused by battery failure or failure,and ensure the normal and stable operation of electronic equipment and system.In this paper,the RUL prediction of lithium-ion battery in the cycle working process is studiedFirst of all,this paper theoretically analyzes the causes of battery aging from the working principle and aging mechanism of lithium-ion battery,and determines that the state of Health(SOH)of battery is the key index to determine the RUL of battery.Theoretical analysis and experimental verification of particle filter algorithm in battery RUL prediction are carried out.In this process,the empirical aging model of battery is studied,and the single exponential empirical aging model is selected as the research focus of this paper..Secondly,the causes of the errors in the experimental results are analyzed.Particle degradation and singular value are the main factors that affect the prediction accuracy of standard particle filter algorithm.In order to ensure the diversity of particles,unscented Kalman filter(UKF)is introduced to optimize the particle filter algorithm to predict RUL.Before the sampling of particle filter algorithm,the UKF is used to re weight the particles in the particle set,so as to ensure the diversity of particles,overcome the problem that the particle filter algorithm will be affected by particle degradation and singular value,and improve the prediction accuracy.Finally,the quantitative expression of the probability density distribution function(PDF)of the prediction results is realized,which provides more valuable information for the maintenance of electronic equipment and system.This paper evaluates the optimization algorithm proposed in this paper by using the relevant indexes widely used in the field of diagnostics health management(PHM)to evaluate the performance of RUL prediction algorithm,and proves that the prediction performance of this algorithm meets the requirements.