Prediction of Pre-Failure States of Lithium-Ion Batteries Based on Cubic Spline Interpolation

Abstract: Failures in lithium-ion batteries (LIBs) frequently occur during operation, making the prediction of their pre-failure state a cru-cial task for preventing emergency situations. In this study, an algorithm for predicting the pre-failure state of LIBs based on the cubic spline method is developed. This method makes it possible to predict a future voltage value in the battery followed by correction by the voltage coming from the sensor. The algo-rithm creates a digital voltage “portrait” of the battery in its pre-failure state that identifies characteristic features that are inter-preted as precursors of impending failure. The efficiency of the proposed approach was demonstrated via a computational task that used the algorithm to identify failure precursors. The con-ducted research demonstrates that the proposed algorithm en-sures a high accuracy of such prediction, providing timely warnings about a potential transition of the system to an emer-gency state.

Keywords: failure precursor, pre-failure state prediction, cubic spline interpolation, lithium-ion battery