Analysis of Li-ion battery degradation using self-organizing maps

Abstract

This paper proposes a new methodology to identify the different degradation processes of Li-Ion battery cells. The goal of this study is to determine if different degradation factors can be separated by waveform analysis from aged cells with similar remaining capacity. In contrast to other works, the proposed method identifies the past operating conditions in the cell, regardless of the actual State of Health. The methodology is based on a data-driven approach by using a SOM (Self-organizing map), an unsupervised neural network. To verify the hypothesis a SOM has been trained with laboratory data from whole data cycles, to classify cells concerning their degradation path and according to their discharge voltage patterns. Additionally, this new methodology based on the SOM allows discriminating groups of cells with different cycling conditions (based on depth of discharge, ambient temperature and discharge current).This research line is very promising for classification of used cells, not only depending on their current static parameters (capacity, impedance), but also the battery use in their past life. This will allow making predictions of the Remaining Useful Life (RUL) of a battery with greater precision.

Unsupervised Neural Networks for Identification of Aging Conditions in Li-Ion Batteries

“…A first version of this methodology was previously explored in [30]. However, the results were not as successful as expected due to the limited dataset used and the lack of a feature extraction stage.…”