Flexible Sensors for Battery Health Monitoring

Abstract

With the widespread application of lithium batteries in electric vehicles and energy storage systems, battery-related safety and reliability issues have become increasingly prominent. Conventional monitoring methods often struggle to address dynamic changes under complex operando. In recent years, flexible sensing technology has emerged as a promising solution for battery health monitoring due to its high adaptability and conformability to complex structures. Meanwhile, empowered by artificial intelligence (AI) for data analysis, the collected data enables efficient and accurate state assessment, offering robust support for accident prevention. Against this background, this paper first explores the integrated applications of flexible sensors in battery health monitoring and their unique advantages in addressing complex battery operating conditions, while analyzing the potential of AI in battery state analysis. Subsequently, it systematically reviews mainstream flexible sensing technologies (e.g., film sensors, thermocouples, and optical fiber sensors), elucidating their mechanisms for revealing intricate internal battery processes during operation. Finally, the paper discusses AI’s role in enhancing monitoring efficiency and accuracy, and envisions future research directions and application prospects. This work aims to provide technical references for the battery health monitoring field as well as promote the application of flexible sensing technologies in improving battery system safety and reliability.

1. Introduction

Lithium-ion batteries are the core components of electric vehicles and scaled energy storage systems [1–3]. The safety and health of these batteries directly determine the reliability and lifetime of energy systems [4–7]. Despite the substantial enhancement in energy density and cycling performance of batteries in recent years [8–11], under complex operando (e.g., mechanical abuse, thermal abuse, or electrical abuse), the coupling failure of multiple physical fields (mechanical, thermal, and electrochemical) within the battery may still trigger catastrophic events such as cascading thermal runaway or even explosion [12–15], resulting in serious safety hazards.

The demand for battery testing has led to significant advancements in non-in situ and in situ techniques over the past few decades [16–18], with these techniques becoming increasingly important in the design of batteries [19, 20]. Nevertheless, the parameters that can be extracted by non-in situ and in situ techniques are frequently detached from the real operating state, i.e., the working environment. Consequently, battery operando detection technology has become a hotspot and a challenge in battery research in recent years [21–25].

Flexible sensing technology provides a breakthrough solution to this challenge by virtue of its thinness (thickness can be as low as micron level) [26–30], high ductility, and low invasiveness. It is capable of real-time monitoring of multi-physical field states under complex battery operando [31–36], and provides real-time feedback to the control system through the parsing system (Fig. 1). The in-depth integration of AI technology enables the monitoring system to extract multi-physical field correlation features (e.g., pressure-internal resistance coupling) from the massive data of the battery operando and establish the complex relationship between these key features and the battery performance to enhance the prediction and synergistic capability of the system [37–42]. For example, in a typical logic closed loop, the electrical, temperature, and pressure signals of the battery system operation acquired by the sensors are synchronously analyzed.