| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Aluminum Reduction Technology
|
| Presentation Title |
AlF3 Shots Prediction for Optimal Temperature Control and Process Efficiency in Aluminium Smelter |
| Author(s) |
Manish Jaiswal, Himan Kundu, Shanmukh Rajgire, Anish Das |
| On-Site Speaker (Planned) |
Manish Jaiswal |
| Abstract Scope |
AlF3 feeding is crucial for maintaining pot thermal balance & electrolyte temperature. The current reactive approach lacks precision, often causing additional feeding and in some cases lesser feeding due to certain factors like misattributed temperature rises, errors in bath chemistry and few operational inefficiencies. This improper thermal control negatively impacts pot performance.
This article presents a predictive model for optimizing AlF3 shots in aluminum smelting operations, leveraging Machine Learning techniques specifically through Support Vector Regression (SVR). Using data from a 360kA pot-line, the model predicts AlF3 shots for 360 pots across different pot-age groups—low, intermediate, mid, and old—ensuring precise and tailored AlF3 feeding strategies. The SVR-based model forecasts AlF3 requirements by analysing key performance indicators and operational metrics, focusing on minimizing Mean Absolute Percentage Error (MAPE).
This predictive model aids in maintaining pot temperature and controlling excess AlF3 with minimal deviation, thereby enhancing stability and improving current efficiency. |
| Proceedings Inclusion? |
Planned: Light Metals |
| Keywords |
Aluminum, Machine Learning, Other |