Performance Comparison of Forecasting the State of Health of Lithium-Ion Battery  
Author Lei Qin

 
Co-Author(s) Ben-Chang Shia;  Mingchih Chen; Jian Li

 
Abstract Machine learning methods are widely used to predict the state of health (SOH) of lithium-ion batteries, but usually have two limitations: (1) the parameters of the prediction models are difficult to determine; (2) most current models are for a single battery. To address this challenge, this paper employs the automachine learning algorithm to perform SOH prediction on experimental data (NASA dataset) and compares the performance with the results by using the traditional linear regression method. In this paper, two machine learning methods, automatic extreme gradient boosting (AutoXgboost), automatic deep neural network (H2oDeepLearning), are used to select their parameters in an automated data-driven manner. The experimental results show that the linear regression method maintains good stability in SOH prediction compared with new machine learning algorithm.

 
Keywords Lithium-ion battery; State of health; machine learning method, linear regression
   
    Article #:  DSBFI23-1
 
Proceedings of 2nd ISSAT International Conference on Data Science in Business, Finance and Industry
January 8-10, 2023 - Da Nang, Vietnam