Open Access

ARTICLE

SOH Estimation of Lithium Batteries Based on ICA and WOA-RBF Algorithm

Qi Wang^1,2,3, Yandong Gu^1,*, Tao Zhu¹, Lantian Ge¹, Yibo Huang¹

1 School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou, 213001, China
2 Research and Development Department, Shuangdeng Group Co., Ltd., Taizhou, 225500, China
3 School of Automotive and Traffic Engineering, Jiangsu University, Zhenjiang, 212013, China

* Corresponding Author: Yandong Gu. Email:

(This article belongs to the Special Issue: Advanced Modelling, Operation, Management and Diagnosis of Lithium Batteries)

Energy Engineering 2024, 121(11), 3221-3239. https://doi.org/10.32604/ee.2024.053758

Received 09 May 2024; Accepted 14 August 2024; Issue published 21 October 2024

Abstract

Accurately estimating the State of Health (SOH) of batteries is of great significance for the stable operation and safety of lithium batteries. This article proposes a method based on the combination of Capacity Incremental Curve Analysis (ICA) and Whale Optimization Algorithm-Radial Basis Function (WOA-RBF) neural network algorithm to address the issues of low accuracy and slow convergence speed in estimating State of Health of batteries. Firstly, preprocess the battery data to obtain the real battery SOH curve and Capacity-Voltage (Q-V) curve, convert the Q-V curve into an IC curve and denoise it, analyze the parameters in the IC curve that may serve as health features; Then, extract the constant current charging time of the battery and the horizontal and vertical coordinates of the two IC peaks as health features, and perform correlation analysis using Pearson correlation coefficient method; Finally, the WOA-RBF algorithm was used to estimate the battery SOH, and the training results of LSTM, RBF, and PSO-RBF algorithms were compared. The conclusion was drawn that the WOA-RBF algorithm has high accuracy, fast convergence speed, and the best linearity in estimating SOH. The absolute error of its SOH estimation can be controlled within 1%, and the relative error can be controlled within 2%.

---

Keywords

---