Non-destructive battery health monitoring based on ultrasound technique and Kalman filtering algorithm

Abstract

Batteries are widely used in various applications, including electric vehicles, renewable energy systems, and portable electronics. Their performance and safety can be affected by various internal and external factors, such as aging, abuse, and defects. The assessment of battery health status is a crucial area of research, which typically involves the use of two primary indicators, namely state-of-charge (SoC) and state-of-health (SoH). These parameters are commonly employed to evaluate the remaining capacity and the remaining life of Li-ion batteries, respectively. While various techniques exist for assessing the SoC and SoH of Li-ion batteries, these methods primarily rely on the acquisition of current and voltage measurements to facilitate battery modeling and lifetime estimation. However, in instances where there is internal damage to the battery, and current and voltage data acquisition is no longer feasible, there is currently no viable system for monitoring battery health. Ultrasonic testing (UT) methods have emerged as valuable tools for monitoring battery health without causing damage or altering their characteristics. In this study, we establish and develop a battery health monitoring system based on ultrasonic transducers, enabling real-time monitoring of ultrasonic signal changes during battery charging and discharging processes. This paper makes several significant contributions. Firstly, it provides novel insights into the variation of ultrasonic signal strength and time-of-flight (ToF) during battery charging and discharging, shedding light on previously unexplored aspects of battery behavior. Secondly, we propose a model for estimating the SoC of rechargeable batteries using ultrasonic echoes and the extended Kalman filter (EKF) method. This innovative approach leverages the correlation between battery SoC and ultrasonic echo characteristics, such as amplitude and ToF, measured at different discharge rates. By employing the EKF algorithm to model system dynamics and estimate state variables, our method surpasses conventional EKF techniques in terms of accuracy and reliability, as demonstrated by comparative analysis. Moreover, our approach proves robust to variations in echo signals, further enhancing its utility for battery health monitoring. Finally, we introduce the use of ultrasonic techniques for detecting overcharging phenomena and bubble generation in batteries, opening new avenues for enhanced battery management systems in electric vehicles, renewable energy storage, and other battery-operated devices. Through these contributions, our study aims to advance the field of battery health monitoring and pave the way for safer and more efficient battery utilization across various applications.