Health State Estimation of Lithium-Ion Battery Based on Equal Time Interval Charging
Received:August 04, 2018  Revised:December 25, 2018
View Full Text  View/Add Comment  Download reader
DOI:10.7643/ issn.1672-9242.2018.12.012
KeyWord:lithium-ion battery  state-of-health  Gaussian process regression  health indicator
        
AuthorInstitution
LIN Tian-tian Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Key Laboratory of Ocean Engineering of Shanghai Jiao Tong University, Shanghai , China
CHEN Zi-qiang Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Key Laboratory of Ocean Engineering of Shanghai Jiao Tong University, Shanghai , China
LIU Jian Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Key Laboratory of Ocean Engineering of Shanghai Jiao Tong University, Shanghai , China
Hits:
Download times:
Abstract:
      Objective To put forward a kind of SOH estimation method based on the health factors, to accurately estimate the health status of lithium ion batteries. Methods The time interval between two constant voltages during the constant current charging process was selected as the health indicator to estimate SOH. The Gaussian process regression method was used to estimate SOH. The hyper-parameter was optimized by the conjugate gradient method. The health indicator was taken as the input of the model to output the corresponding SOH. The experimental data of six batteries under different experimental conditions from NASA was selected to verify the method. Results The MAPE and RMSE values of the estimated results of the 6 batteries selected were all below 0.02. Conclusions The health indicator selected can better characterize the SOH of the battery. It verifies the feasibility of the SOH estimation method based on health indicator. The method can accurately estimate the SOH of batteries under different temperatures, discharge rates and depths of discharge, and has strong applicability.
Close