Advanced online method for battery model parameter identification: Bias-compensated forgetting factor recursive least squares

Lithium-ion power battery technology stands out as a pivotal component in advancement of new energy electric vehicles (EVs). Battery parameter identification, as one of the core technologies to achieve an efficient battery management system (BMS), is the key to predicting and managing the performance of Li-ion batteries. A recent breakthrough study presented by researchers from Hebei University of Technology proposes an online battery model parameters identification approach based on bias-compensated forgetting factor recursive least squares. This advanced method is expected to improve the accuracy of parameter identification under different noise.

The essence of battery parameter identification lies in choosing the accurate lithium battery model and selecting an appropriate model parameter identification method. For this research, a Bias-Compensated forgetting factor recursive least squares (BCFFRLS) method based on bias compensation is proposed for application in dual-polarized equivalent circuit models. It can find the noise mean-square deviation of the signal contamination by constructing a generalization matrix when both input and output are contaminated with noise.

In dynamic and complex operational scenarios, the presence of randomly sampled noise interferes with measurements of voltage and current, compromising accuracy of parameter identification for battery model. The BCFFRLS method performs well under various complex operating conditions. Comparative analysis reveals substantial improvements, with the mean absolute error reduced by 25%, 28%, and 15%, and the root mean square error reduced by 25.1%, 42.7%, and 15.9% in Urban Dynamometer Driving Schedule (UDDS), Dynamic Stress Test (DST), and Hybrid Pulse Power Characterization (HPPC) operating conditions, respectively, when compared to the Forgetting Factor Recursive Least Squares (FFRLS) method.

The BCFFRLS method shows that BCFFRLS algorithm has some improvements in computation times compared to FFRLS, and it has moderate computation which can also be used for online identification. As online identification technology matures, it may drive further innovation in battery technology, fostering advancements in the energy sector. It could also stimulate the development of related industries, such as high-precision sensors, data analysis algorithms, and intelligent control systems.

The BCFFRLS method mainly improves the inaccuracy of model parameter estimation when the real values of current and voltage are contaminated by white noise. In the future, how to design battery parameter identification models in case of sudden failure of current and voltage acquisition.

Reference

[1] He H, Sun F, Wang Z, Lin C, Zhang C, Xiong R, et al. China’s battery electric vehicles lead the world: achievements in technology system architecture and technological breakthroughs. Green Energy Intellig Transp 2022;1(1).

[2] Michaelides EE, Nguyen VND, Michaelides DN. The effect of electric vehicle energy storage on the transition to renewable energy. Green Energy Intellig Transp 2023;2(1).

[3] Chen C, Xiong R, Yang R, Li H. A novel data-driven method for mining battery open-circuit voltage characterization. Green Energy Intellig Transp 2022;1(1).

[4] Zheng Yuejiu, Ouyang Minggao, Han Xuebing, Lu Languang, Li Jiangqiu. Investigating the error sources of the online state of charge estimation methods for lithium-ion batteries in electric vehicles. J Power Sources 2018;377:161–88.

Author: Dong Zhen, Jiahao Liu, Shuqin Ma, Jingyu Zhu, Jinzhen Kong, Yizhao Gao, Guojin Feng, Fengshou Gu

Title of original paper: Online battery model parameters identification approach based on bias-compensated forgetting factor recursive least squares

Article link: https://doi.org/10.1016/j.geits.2024.100207

Journal: Green Energy and Intelligent Transportation

https://www.sciencedirect.com/science/article/pii/S2773153724000598