Quantitative Prediction of Coal–Gangue Content Using Terahertz Time-Domain Spectroscopy and Physics-Informed Machine Learning
Zeping Liu, Lipeng Hu, Jianfei Xu, Yadong Yang, Sitong Li, Zhou Xu, Longhai Liu, Jiabao Li, Houli Liu, Dongdong Ye · 2026
Read Quantitative Prediction of Coal–Gangue Content Using Terahertz Time-Domain Spectroscopy and Physics-Informed Machine Learning by Zeping Liu, Lipeng Hu, Jianfei Xu, Yadong Yang, Sitong Li, Zhou Xu, Longhai Liu, Jiabao Li, Houli Liu, Dongdong Ye (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13567379.txt.
Description
Zeping Liu and nine co-authors published a spectroscopy study in twenty twenty-six on measuring how much worthless rock hides in coal, using terahertz waves and physics-informed machine learning. Knowing gangue content guides coal use and automated sorting. Terahertz scans captured time and frequency signals plus optical constants from mixtures of known composition.
Five algorithms competed under cross-validation. A neural network constrained by effective medium theory, forced to respect the physics of mixtures, won clearly. Its predictions explained most variance with errors of a few percentage points, beating random forests, support vector, Gaussian process, and plain neural rivals. The authors present the hybrid as fast, interpretable quality control for an industry that still samples by hand.
Catalog record
| CREATOR | Zeping Liu, Lipeng Hu, Jianfei Xu, Yadong Yang, Sitong Li, Zhou Xu, Longhai Liu, Jiabao Li, Houli Liu, Dongdong Ye |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |