Scholarlink

<p>The emergence and rapid advancement of artificial intelligence (AI) mark the beginning of a transformative era in technology and science. This new era seems to affect classical thermal analysis approaches. This study broadly explores the integration of machine learning approaches within thermal analysis applications. AI transforms the field through three main paradigms: predictive modeling, enhanced data interpretation using labeled data, and optimization and kinetic analysis. The paper discusses the implementation of various machine learning architectures, challenges in their application, and their role in thermal dehydration and lifetime prediction. Notably, Physics-Informed Neural Networks (PINNs) enhance physical and thermodynamic consistency while reducing experimental effort, and machine learning models show strong potential for thermal degradation and lifetime estimation of materials. This study provides a comprehensive framework for understanding current capabilities, identifies critical research gaps, and charts paths toward intelligent, adaptive thermal analysis systems and discusses the future of the topic.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The place and future of AI in thermal analysis processes: from prediction to foresight

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月15日

查看全文
  • Emre Arpaci

摘要

The emergence and rapid advancement of artificial intelligence (AI) mark the beginning of a transformative era in technology and science. This new era seems to affect classical thermal analysis approaches. This study broadly explores the integration of machine learning approaches within thermal analysis applications. AI transforms the field through three main paradigms: predictive modeling, enhanced data interpretation using labeled data, and optimization and kinetic analysis. The paper discusses the implementation of various machine learning architectures, challenges in their application, and their role in thermal dehydration and lifetime prediction. Notably, Physics-Informed Neural Networks (PINNs) enhance physical and thermodynamic consistency while reducing experimental effort, and machine learning models show strong potential for thermal degradation and lifetime estimation of materials. This study provides a comprehensive framework for understanding current capabilities, identifies critical research gaps, and charts paths toward intelligent, adaptive thermal analysis systems and discusses the future of the topic.