海归学者发起的公益学术平台
分享信息，整合资源
交流学术，偶尔风月
锂离子电池（LIBs）在支持设备电气化方面至关重要，但优化未来的电池仍面临挑战。其关键问题在于如何快速识别电池退化原因及防止容量过早衰减。由于非线性容量损失可能发生在数百或数千个周期内，探索电池材料退化机制费时耗力。另外，早期电池寿命预测的另一个挑战是阳极、阴极和电解质中电池化学成分的多样性，以及形状因子和电化学测试协议的多样性。机器学习（ML）和深度学习（DL）可以通过减少理解基本化学过程所需的循环次数来加速测试。但这些模型通常专注于单任务学习，而忽视多目标学习。它们往往忽略生产差异和个体电池差异，这使它们具有较低的精度和可信度。
Fig. 1 | An UML diagram of the computational framework.
来自德国亥姆霍兹研究所的Fuzhan Rahmanian等人，展示了一种不依赖电池化学成分、形式和循环程序的ARCANA框架，通过多任务学习和注意力机制，来可靠地监测电池寿命和健康状态的能力。该模型在每个循环中整合了不确定性量化和注意力机制，以明确每次预测中的模型关注点，这对于发现与多个因素关联的复杂模式至关重要。ARCANA的模块化设计支持电池退化的实时监测，促进及时和成本效益高的干预，进而改进研发过程，并为材料选择和协议优化提供信息。通过自动化数据收集、处理和分析，研究人员可以简化实验流程，减少人为错误。机器学习模型可以不断从新数据中学习，适应不断变化的实验条件，提供实时见解。
Fig. 2 | Comparative analysis of model predictions and its uncertainty and calibration for Qdis in cylindrical sample cells.
这一整合有潜力彻底改变电池研究，使研究人员能更快做出数据驱动决策，加快发现速度。该文近期发表于npj Computational Materials 10:100 2024，英文标题与摘要如下，点击左下角“阅读原文”可以自由获取论文PDF。
Attention towards chemistry agnostic and explainable battery lifetime prediction
Fuzhan Rahmanian, Robert M. Lee, Dominik Linzner, Kathrin Michel, Leon Merker, Balazs B. Berkes, Leah Nuss & Helge Sören Stein
Predicting and monitoring battery life early and across chemistries is a significant challenge due to the plethora of degradation paths, form factors, and electrochemical testing protocols. Existing models typically translate poorly across different electrode, electrolyte, and additive materials, mostly require a fixed number of cycles, and are limited to a single discharge protocol. Here, an attention-based recurrent algorithm for neural analysis (ARCANA) architecture is developed and trained on an ultra-large, proprietary dataset from BASF and a large Li-ion dataset gathered from literature across the globe. ARCANA generalizes well across this diverse set of chemistries, electrolyte formulations, battery designs, and cycling protocols and thus allows for an extraction of data-driven knowledge of the degradation mechanisms. The model’s adaptability is further demonstrated through fine-tuning on Na-ion batteries. ARCANA advances the frontier of large-scale time series models in analytical chemistry beyond textual data and holds the potential to significantly accelerate discovery-oriented battery research endeavors.
扩展阅读
