Artificial Intelligence Analysis of State of Charge Distribution in Lithium-Ion Battery Based on Ultrasonic Scanning Data

Lithium-ion batteries are the most prevelant electrochemical energy storage devices, but they often suffer from inconsistent charging/discharging speeds at different positions in the cell. This problem would lead to uneven distribution of state of charge (SOC), and may cause capacity degradation acceleration. To address this issue, this paper proposes a novel non-destructive method for characterizing the distribution of SOC within the battery: by using ultrasonic C-scan technology to collect ultrasonic transmission waveforms at different positions inside the battery, establishing the correlation between battery SOC and ultrasonic waveforms using convolutional neural networks, and further analyzing the non-uniformity of the ultrasound signals to infer the SOC differences at different locations within the battery. The research findings in this paper provide valuable insights for understanding the failure mechanisms of lithium-ion batteries and guiding battery fabrication process optimization.