Abstract No.C180131-169
Author name(s): Jian LIU, Ziqiang CHEN, Deyang HUANG, Shiyao ZHOU, Yu JIANG, Tiantian LIN
Company: State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, China
Accurate prediction of the state of health (SOH) of lithium-ion battery is of great significance for safe operation of electric ships and unmanned underwater vehicles. A new method is proposed with a novel health indicator (HI) and an optimized Gaussian process regression (GPR) model to deal with the difficulties in measuring capacity directly and online predicting SOH accurately. Firstly, the time interval of equal charging voltage difference (TI_ECVD) is extracted as HI which can be adopted as an effective signature of lithium-ion batteries’ health state. Then, a generalized linear regression model is used to analyze the relationship between SOH and TI_ECVD. The optimized GPR model is established with combined kernel functions and particle swarm optimization (PSO) algorithm in order to improve the generalization ability and prediction accuracy. The optimized GPR model is utilized to train the lithium-ion batteries’ charge and discharge cycles as input for GPR model and HI as its output. After the optimized GPR model outputting prediction results of TI_ECVD, accurate prediction of SOH can be accomplished through the generalized linear regression model. Furthermore, the verification experiments are conducted by using the data sets of charge and discharge tests of lithium-ion batteries. The results show that the proposed approach can predict nonlinear degradation of SOH well and have high online prediction accuracy and adaptability. Hence, the presented method in the paper can potentially be used to online predict SOH of lithium-ion batteries for prognostic and health management of electric vessels and undersea robots.
KEY WORDS: electric ship; lithium-ion battery; state of health; gaussian process regression; health indicator
All Rights Reserved. If you need the entire text, please contact the author or intl.exchange@ssnaoe.org