Ultrasonic guided waves as an indicator for the state-of-charge of Li-ion batteries

Data
This dataset forms part of the data used for the following article https://doi.org/10.1016/j.jpowsour.2023.233189. It includes the ultrasonic guided wave data collected during one 0.34C discharge-charge cycle recorded on 13th July 2022. The sensor set-up and required post processing is detailed in the related paper. The data is maintained in its original record format (oscilloscope readings). The following data is included:
• Chirp input signal (1 mat file)
• Responses of cell to input chirp signal throughout discharge-charge cycle (104 mat files)
• Text file containing the voltage, current, temperature, and other parameters.

Signal type
Generally, a single tone burst signal may be used to excite a cell (e.g., fig 2 e, blue). Correspondingly, the response to this specific tone burst signal with one specific central frequency (fig 2 g, blue) could be measured. However, when collecting these ultrasonic measurements, it wasn’t decided yet which parameters (center frequency, number of excitation cycles) would be ideal for the probed cell. It was therefore decided to excite the cell with chirp signals. A chirp signal contains a wide range of frequencies (fig 2 b). Therefore, the response signal is chaotic (fig 2 c). It’s therefore hard to analyze the response to a chirp signal directly. Using the chirp signal and the response to the chirp signal the transfer function of the cell at each measurement point was calculated and used to compute the expected response to a tone burst signal (formula 4 / fig 2 e and g, red). This allows to quickly try different center frequencies in the post processing and analyze which frequency is best suited for a specific cell (see figure 3).

Lab protocol
The mat files of the ultrasonic measurements include measurements from all paths between the four sensors (see figure 1). For the analysis in figure 6 the data from sensor 1 (emitter) to sensor 2 (receiver) has been used. A pdf explaining the structure of the mat files is attached. The ultrasonic probing was done in time intervals of 5 minutes not SoC intervals. The SoC can be calculated based on the charge drawn from the cell which is included in the electrical measurements. The ultrasonic chirp signal data and the extrinsic cell parameters may be correlated using the time stamp of each recording. The following steps are included in the data.

13.07.2022: Charge/Discharged cell “Palma” (0.34C)
1. Charge at 6.25A to 4.2 V
2. Charge at 4.2V until I<0.625A
3. Rest 30 min
4. Discharged at 4.25A to 3V
5. Discharged at 3V until I<0.625A
6. Rest 30 min
7. Charge at 4.25A to 4.2 V
8. Charged at 4.2V until I<0.625A
9. Rest 30 min
10. Discharged 2500mAh at 12.5A
Between steps 1-9 the following SHM measurement were taken automated every 5 min
A. Chirp_24V_100MHz_100us_1-600kHz (Palma20220713Chirp_i)

数据集
本数据集为下述论文所用数据的一部分：https://doi.org/10.1016/j.jpowsour.2023.233189。其包含2022年7月13日采集的一次0.34C倍率充放电循环过程中的超声导波数据。传感器布设方案与所需的后处理流程详见相关研究论文。数据保留原始记录形式，即示波器采集的原始读数。本次数据集包含以下内容：
• 线性调频脉冲（Chirp）输入信号（1个MAT文件）
• 充放电循环全程中电池对输入Chirp信号的响应数据（104个MAT文件）
• 包含电压、电流、温度及其他参数的文本文件。

信号类型
通常可采用单音猝发信号激励电池（如图2e蓝色曲线所示），并同步采集该特定中心频率下单音猝发信号对应的电池响应（如图2g蓝色曲线所示）。但在超声测量阶段，尚未确定适配待测电池的最优参数（中心频率、激励循环次数等），因此最终采用Chirp信号对电池进行激励。Chirp信号覆盖宽频率范围（如图2b所示），其对应的电池响应信号呈混沌特性（如图2c所示），难以直接开展响应分析。通过结合Chirp输入信号与对应电池响应，可计算得到各测量点处电池的传递函数，并据此求解单音猝发信号的理论响应（详见公式4、图2e与2g红色曲线）。该方法可在后处理阶段快速测试不同中心频率的适配效果，进而分析针对特定电池的最优激励频率（详见图3）。

实验室实验规程
超声测量的MAT文件包含4个传感器间所有检测路径的测量数据（详见图1）。针对图6的分析，本数据集采用了传感器1（发射端）至传感器2（接收端）的路径数据。附带的PDF文件详细说明了MAT文件的结构。本次超声探测以5分钟为时间间隔进行采样，而非按荷电状态（State of Charge, SoC）间隔采样。可根据电气测量数据中记录的电池放电电荷量计算SoC。通过各条记录的时间戳，可实现超声Chirp信号数据与电池外部参数的关联。本数据集包含以下实验流程：

2022年7月13日：对编号为"Palma"的电池开展0.34C倍率充放电测试
1. 以6.25A恒流充电至4.2V
2. 以4.2V恒压充电至电流小于0.625A
3. 静置30分钟
4. 以4.25A恒流放电至3V
5. 以3V恒压放电至电流小于0.625A
6. 静置30分钟
7. 以4.25A恒流充电至4.2V
8. 以4.2V恒压充电至电流小于0.625A
9. 静置30分钟
10. 以12.5A恒流放电2500mAh
在步骤1至9之间，每5分钟自动采集一次结构健康监测（Structural Health Monitoring, SHM）数据：
A. 采用Chirp_24V_100MHz_100us_1-600kHz信号进行测试（对应命名为Palma20220713Chirp_i）