| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Alumina and Bauxite
|
| Presentation Title |
Combine LIBS and Machine Learning to Realize Real-Time and Online Analysis of Bauxite Composition |
| Author(s) |
Yanfang Zhang, Lijuan Qi, Long Duan, Qiaoyun Liu, Jun Wu, Shuai Shao, Hui Pan |
| On-Site Speaker (Planned) |
Yanfang Zhang |
| Abstract Scope |
The bauxite composition determines the material ratio of raw slurry preparation process in alumina production. The real-time and online analysis of bauxite composition can guide the material ratio and reduce the fluctuation of process indicators in alumina production. In this study, based on laser-induced breakdown spectroscopy (LIBS) technology, collected the spectral data of bauxite, preprocessed the spectral data, selected feature wavelength, and constructed the modeling dataset of element analysis. Combined with technics principles driven and data driven, established the analysis model by machine learning. The accuracy of content analysis of Al, Fe and Si were greater than 90%. The results showed that the real-time and online analysis of bauxite composition was realized by combining LIBS and machine learning technology, which laid the foundation for the realization of intelligent control of charge mixture, and accelerated the information and intelligence process of alumina production process. |
| Proceedings Inclusion? |
Planned: Light Metals |
| Keywords |
Other, Other, Other |