Zehua Wang, Li Wang, Hao Zhang, Hong Xu, Xiangming He
{"title":"数据驱动法中精确、合适的描述符对促进锂电池开发的重要性:透视","authors":"Zehua Wang, Li Wang, Hao Zhang, Hong Xu, Xiangming He","doi":"10.1002/elt2.41","DOIUrl":null,"url":null,"abstract":"<p>Conventional approaches for developing new materials may no longer be adequate to meet the urgent needs of humanity's energy transition. The emergence of machine learning (ML) and artificial intelligence (AI) has led materials scientists to recognize the potential of using AI/ML to accelerate the creation of new battery materials. Although fixed material properties have been extensively studied as descriptors to establish the link between AI and materials chemistry, they often lack versatility and accuracy due to a lack of understanding of the underlying mechanisms of AI/ML. Therefore, materials scientists need to have a comprehensive understanding of the operational mechanisms and learning logic of AI/ML to design more accurate descriptors. This paper provides a review of previous research studies conducted on AI, ML, and descriptors, which have been used to address challenges at various levels, ranging from materials development to battery performance prediction. Additionally, it introduces the basics of AI and ML to assist materials and battery developers in comprehending their operational mechanisms. The paper demonstrates the significance of precise and suitable ML descriptors in the creation of new battery materials. It does so by providing examples, summarizing current descriptors and ML algorithms, and examining the potential implications of future AI advancements for the sustainable energy industry.</p>","PeriodicalId":100403,"journal":{"name":"Electron","volume":"2 4","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/elt2.41","citationCount":"0","resultStr":"{\"title\":\"The importance of precise and suitable descriptors in data-driven approach to boost development of lithium batteries: A perspective\",\"authors\":\"Zehua Wang, Li Wang, Hao Zhang, Hong Xu, Xiangming He\",\"doi\":\"10.1002/elt2.41\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Conventional approaches for developing new materials may no longer be adequate to meet the urgent needs of humanity's energy transition. The emergence of machine learning (ML) and artificial intelligence (AI) has led materials scientists to recognize the potential of using AI/ML to accelerate the creation of new battery materials. Although fixed material properties have been extensively studied as descriptors to establish the link between AI and materials chemistry, they often lack versatility and accuracy due to a lack of understanding of the underlying mechanisms of AI/ML. Therefore, materials scientists need to have a comprehensive understanding of the operational mechanisms and learning logic of AI/ML to design more accurate descriptors. This paper provides a review of previous research studies conducted on AI, ML, and descriptors, which have been used to address challenges at various levels, ranging from materials development to battery performance prediction. Additionally, it introduces the basics of AI and ML to assist materials and battery developers in comprehending their operational mechanisms. The paper demonstrates the significance of precise and suitable ML descriptors in the creation of new battery materials. It does so by providing examples, summarizing current descriptors and ML algorithms, and examining the potential implications of future AI advancements for the sustainable energy industry.</p>\",\"PeriodicalId\":100403,\"journal\":{\"name\":\"Electron\",\"volume\":\"2 4\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-06-02\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://onlinelibrary.wiley.com/doi/epdf/10.1002/elt2.41\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Electron\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/elt2.41\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Electron","FirstCategoryId":"1085","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/elt2.41","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
开发新材料的传统方法可能已无法满足人类能源转型的迫切需求。机器学习(ML)和人工智能(AI)的出现使材料科学家认识到利用 AI/ML 加速创造新型电池材料的潜力。虽然固定的材料属性已被广泛研究,作为建立人工智能与材料化学之间联系的描述符,但由于缺乏对人工智能/ML 潜在机制的了解,这些属性往往缺乏通用性和准确性。因此,材料科学家需要全面了解人工智能/ML 的运行机制和学习逻辑,以设计出更精确的描述符。本文回顾了以往关于人工智能、ML 和描述符的研究,这些研究已被用于应对从材料开发到电池性能预测等不同层面的挑战。此外,论文还介绍了人工智能和 ML 的基础知识,以帮助材料和电池开发人员理解其运行机制。本文展示了精确、合适的 ML 描述符对创造新型电池材料的重要意义。本文通过举例说明、总结当前的描述符和 ML 算法,以及研究未来人工智能进步对可持续能源行业的潜在影响来实现这一目的。
The importance of precise and suitable descriptors in data-driven approach to boost development of lithium batteries: A perspective
Conventional approaches for developing new materials may no longer be adequate to meet the urgent needs of humanity's energy transition. The emergence of machine learning (ML) and artificial intelligence (AI) has led materials scientists to recognize the potential of using AI/ML to accelerate the creation of new battery materials. Although fixed material properties have been extensively studied as descriptors to establish the link between AI and materials chemistry, they often lack versatility and accuracy due to a lack of understanding of the underlying mechanisms of AI/ML. Therefore, materials scientists need to have a comprehensive understanding of the operational mechanisms and learning logic of AI/ML to design more accurate descriptors. This paper provides a review of previous research studies conducted on AI, ML, and descriptors, which have been used to address challenges at various levels, ranging from materials development to battery performance prediction. Additionally, it introduces the basics of AI and ML to assist materials and battery developers in comprehending their operational mechanisms. The paper demonstrates the significance of precise and suitable ML descriptors in the creation of new battery materials. It does so by providing examples, summarizing current descriptors and ML algorithms, and examining the potential implications of future AI advancements for the sustainable energy industry.