Congratulations to Yunfei Han and Jia Xie for their joint research paper " Capacity estimation of lithium-ion battery based on Gaussian process regression and feature selection " accepted by Energy Storage Science and Technology

Abstract

Because of the complex degradation characteristics of Li-ion batteries, it is a continuing challenge to accurately predict the state of health and remaining useful life of batteries, which limits the development of consumer electronics, electric vehicles, and grid energy storage technologies. The degradation mechanism of the batteries is complex and coupled with each other. Therefore, it is difficult to use a model-based method for accurate modeling. This study proposes a data-driven capacity estimation method for Li-ion batteries.By analyzing the evolution pattern of the voltage-discharge capacity curve with cycle aging,features with an electrochemical concept are selected as the model input, and the capacity of Li-ion batteries can be predicted by the Gaussian process regression(GPR) model. The input features of the model can be obtained online, and the capacity of the battery can be estimated without a full charge-discharge. The experimental verification was completed with the data sets for LCO/graphite batteries and LFP/graphite batteries. The results show that the method has a good generalization ability and can accurately estimate the capacity of different types of batteries. The proposed method is compared to the GPR model with electrochemical impedance spectroscopy as the input, and the results indicate that the proposed method can obtain better estimation accuracy. This highlights that the appropriate selection of various features can significantly affect the performance of the data-driven model for Li-ion batteries and can provide a reference for battery state prediction and diagnosis.

https://kns.cnki.net/kcms/detail/detail.aspx?doi=10.19799/j.cnki.2095-4239.2021.0109