Adaptive detection of terminal voltage collapses for Li-ion batteries

Abstract

We introduce a novel approach for detecting and predicting terminal voltage collapses in Li-ion batteries without having complete knowledge of a battery model. We present a simplified dynamic model for a Li-ion battery that is forced to track the output voltage curve of a physical Li-ion battery by using universal adaptive stabilization (UAS). We prove that when the physical Li-ion battery becomes unstable, then the simplified dynamic model becomes unstable. Our results do not require a sophisticated model for a battery that faithfully captures all dynamics. Using our results, we present an algorithm for detecting impending voltage collapses for Liion batteries.