高等学校化学学报 ›› 2022, Vol. 43 ›› Issue (6): 20220040.doi: 10.7503/cjcu20220040

• 分析化学 • 上一篇    下一篇

超声衰减谱测量电池浆料的粒度分布

黄明心, 周蕾(), 王学重()   

  1. 北京石油化工学院新材料与化工学院, 恩泽生物质精细化工北京市重点实验室, 北京 102617
  • 收稿日期:2022-01-15 出版日期:2022-06-10 发布日期:2022-03-27
  • 通讯作者: 周蕾,王学重 E-mail:zhoulei2020@bipt.edu.cn;wangxuezhong@bipt.edu.cn
  • 基金资助:
    国家自然科学基金(52102208)

Measurement of Particle Size Distribution of Battery Slurries Using Ultrasonic Attenuation Spectroscopy

HUANG Mingxin, ZHOU Lei(), WANG Xuezhong()   

  1. Beijing Key Laboratory of Enze Biomass and Fine Chemicals,College of New Materials and Chemical Engineering,Beijing Institute of Petrochemical Technology,Beijing 102617,China
  • Received:2022-01-15 Online:2022-06-10 Published:2022-03-27
  • Contact: ZHOU Lei,WANG Xuezhong E-mail:zhoulei2020@bipt.edu.cn;wangxuezhong@bipt.edu.cn
  • Supported by:
    the National Natural Science Foundation of China(52102208)

摘要:

电池浆料中颗粒状活性物质的粒度大小和分散均匀性对电池的内阻、 电压、 局部表面电流和总极化程度等性能有直接影响, 实现对其的在线实时测量对电池的质量控制具有重要意义. 基于电池浆料的高固含量、 高黏度和低透光性的特点, 本文利用超声衰减谱的方式测量了其粒度分布(PSD). 应用于电池浆料的粒度分布测量的最大难点是其利用超声衰减谱法预测粒度分布的模型需要难以获得的分散相和连续相的物性参数. 本文采用主成分分析(PCA)结合误差反向传播(BP)神经网络建立预测模型解决了超声衰减谱法的难点, 并引入遗传算法(GA)优化BP神经网络的初始阈值和权值. 通过以LiCoO2为活性物质的电池浆料进行了验证, 结果表明, PCA-GA-BP神经网络能够有效对不同固含量电池浆料的粒度分布进行预测, 预测值与真实值的峰形重合度高, 峰高偏差小, 两者的均方误差为0.1358, 拟合度(R2)为0.9816, 说明超声衰减谱法可作为测量电池浆料粒度分布的重要方式.

关键词: 超声衰减谱, 粒度分布, 电池浆料, BP神经网络, 主成分分析, 遗传算法

Abstract:

The particle size distribution and dispersion uniformity of the granular active material in the battery slurries have a direct impact on the important properties of the battery such as internal resistance, voltage, local surface current, and total polarization degree. It is of great significance for the quality control of the battery if its present line can be measured in real time. Based on the characteristics of high solid content, high viscosity and low transmittance of battery slurry, this paper investigates the measurement of its particle size distribution by ultrasonic attenuation spectroscopy. The biggest difficulty in applying to the particle size distribution measurement of battery slurries is that its model for predicting the particle size distribution using the ultrasound attenuation spectroscopy requires difficult-to-obtain physical parameters of the dispersed and continuous phases. In this paper, principal component analysis(PCA) combined with error back propagation(BP) neural network is proposed to establish a prediction model to solve the difficulties of ultrasonic attenuation spectroscopy method, and genetic algorithm(GA) is introduced to optimize the initial weights and thresholds of the BP network. Combined with the battery slurry with LiCoO2 as the active material for validation, the results show that the PCA-GA-BP neural network can effectively predict the particle size distribution of battery slurry with different solid contents, and the predicted values have high peak shape overlap with the real values and small peak height deviation, with the mean square error of 0.1358 and the degree of fit(R2) of 0.9816, indicating that ultrasonic attenuation spectroscopy can be used as an important way to measure the particle size distribution of battery slurry.

Key words: Ultrasonic attenuation spectroscopy, Particle size distribution, Battery slurry, BP neural network, Principal component analysis, Genetic algorithm

中图分类号: 

TrendMD: