CDS&E/Collaborative Research: A Symbolic Artificial Intelligence Framework for Discovering Physically Interpretable Constitutive Laws of Soft Functional Composites

这项由计算和数据支持的科学与工程(CDS&amp；E)合作研究基金支持基础研究，以实现使用符号人工智能(AI)自动发现软功能复合材料的本构规律。软功能复合材料对新兴的技术和经济领域至关重要，如可伸缩和可穿戴的电子产品、软机器人以及传感和致动。传统的发现本构法则的方法通常是耗时、低效的，而且对于复杂的案件往往无效。该项目不仅将符号人工智能技术用于软功能复合材料，还将用于其他材料类别，从而显著加快本构规律的发现过程。开发的软件工具将通过一个网站向研究和教育界提供。该项目还将培训本科生和研究生，包括未被充分代表的群体，人工智能、力学和材料科学跨学科领域的劳动力。该项目将建立一个符号人工智能框架，用于发现软功能复合材料的物理可解释本构关系，涉及本构关系的一般张量函数，包括标量和张量算子，以及对噪声数据的统计分析。该项目将使用修改的表示法和对称性理论建立不同各向异性材料的本构关系的一般张量函数。符号人工智能体系结构将建立在基因组表示、物理约束、符号树搜索和统计建模的基础上，并将应用于发现和解释具有各种微观结构的软功能复合材料的力学物理本构关系。该工具和它提供的见解将加快软功能复合材料的设计和应用，并最终加速其他材料类别的设计和应用。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This Computational and Data-Enabled Science and Engineering (CDS&E) collaborative research grant supports fundamental research to enable automatic discovery of constitutive laws for soft functional composites using symbolic artificial intelligence (AI). Soft functional composites are essential to emerging technological and economic areas such as stretchable and wearable electronics, soft robotics, and sensing and actuation. Conventional methods for constitutive law discovery are usually time-consuming, inefficient, and often ineffective for complex cases. This project will significantly accelerate the constitutive law discovery process by leveraging symbolic AI technology not only for soft functional composites but also other material classes. The software tools developed will be made available to the research and education communities through a website. This project will also train undergraduate and graduate students, including underrepresented groups, workforce in the transdisciplinary area of AI, mechanics, and materials science.The project will establish a symbolic AI framework for discovering physically interpretable constitutive laws of soft functional composites, involving general tensor functions of constitutive laws, both scalar and tensor-based operators, and statistical analysis for noisy data. The project will establish general tensor functions of constitutive laws for different anisotropic materials using modified representation and symmetry theories. The symbolic-AI architecture will be built upon genome representation, physics constraints, symbolic tree search, and statistic modeling, and will be applied to discover and interpret mechano-physical constitutive laws of soft functional composites with various microstructures. The tool and the insights it provides will expedite the design and applications of soft functional composites, and eventually other material classes.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.