About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Aluminum Reduction Technology
|
Presentation Title |
ML-Powered Pot Performance Prediction in Aluminium Smelter |
Author(s) |
Manish Jaiswal, Shanmukh Rajgire, Atanu Maity, Kishor Pattnaik, Philip Hansda, Pramod Shukla, Kazi Arshad Ansari, Pratap Sahu |
On-Site Speaker (Planned) |
Manish Jaiswal |
Abstract Scope |
Hirakud Aluminum Smelter, India's one of the oldest smelter began in 1959 with a 55-kA pot line and 10 kTPA capacity. Now, it operates at 176kTPA with four lines: three at 85 kA and one at 235 kA. Facing rising energy costs, aluminium smelters aim to reduce specific energy consumption by enhancing current efficiency (CE), which lowers energy use and boosts productivity.
This article presents a predictive model using Artificial Neural Networks (ANN) to identify low-efficiency pots in 235kA aluminium smelting operations. Utilizing data from a 235kA pot-line, the model predicts pot efficiency with high accuracy and precision by analyzing key performance indicators and operational metrics. By providing real-time insights, it pinpoints underperforming pots, enabling targeted interventions that enhance overall process efficiency and productivity. The findings demonstrate the potential of deep learning in optimizing industrial processes and contribute to significant improvements in operational efficiency and resource utilization. |
Proceedings Inclusion? |
Planned: Light Metals |
Keywords |
Aluminum, Machine Learning, Other |