A Quick Screening Approach Based on Fuzzy C-Means Algorithm for the Second Usage of Retired Lithium-Ion Batteries

With numerous lithium-ion batteries retired from electric vehicles, the studies on the battery second usage are extremely imminent. However, existing screening approaches on plenty of cells fail to guarantee high efficiency and high accuracy simultaneously. This article proposes a quick and accurate screening method based on the improved fuzzy c-means (FCM) algorithm. First, the partial charging curve of every single cell is selected optimally based on the incremental capacity analysis, which is frequently used to detect the battery aging mechanism. Second, four important features are extracted from the partial charging curves, including key point, curve gradient, voltage energy, and volatility. Furthermore, feature optimization is done by observing the relationship between capacity and feature. Finally, the retired batteries are screened with the optimal features using the improved FCM algorithm. The screening result on 176 LiFePO4 batteries proves the high accuracy and high efficiency of the approach. Compared with the support vector machine and neural network approaches, the proposed method has better generality and higher efficiency without pretreatment training. The screening accuracy can reach 90.9%. With a permitted error of 1%, it can be as high as 95.5%. The screening efficiency is about 7.6 times higher than the supervised screening method.

Automated summary

The study proposes a quick and accurate screening method for retired batteries based on an improved fuzzy c-means algorithm, achieving high efficiency and accuracy through feature extraction and optimization of partial charging curves. The approach outperforms support vector machine and neural network methods in generality and efficiency, with a screening accuracy of 90.9% and a potential accuracy of 95.5% with a permitted error of 1%, while the efficiency is about 7.6 times higher than supervised screening methods.