Level 2 Automation of Run-out Table
Cooling the sheet after hot rolling is of great importance, so producing high-quality and valuable sheets requires cooling based on a favorable temperature profile for each steel grade. Therefore, this project was defined in line with the desire for this goal.
Saba steel run-out table includes 19 curtain showers in the upper part of the sheet and 19 point showers in the lower part of the sheet. All these showers have a control loop. Since the amount of water and water vapor on the sheet is extremely high along the entire path of the table, it is not possible to measure the surface temperature of the sheet directly. To solve this problem, the heat transfer model is extracted, and the temperature of the sheet is estimated at all points based on the model with high accuracy.
Along with the online control software, offline software was also designed and provided to the operating engineers to help better analysis.
Purpose of Project
Adjusting the flow rate of the run-out table showers to achieve the desired temperature profile
Steps of Project
- Extracting the data required for modeling from the archive
- Checking the performance of level one
- Derivation of heat transfer model of Saba steel run-out table
- Adding adaptation to the inference model with deep learning methods
- Implementing a suitable controller for extracting the flow rate of showers
- Adding adaptation to the controller
- Adding feedforward to the controller
- Implementation of the model and controllers in the existing communication platform in Saba Steel (Danieli)
Used Technologies
- Heat transfer modeling
- Classic control
- Deep learning
- Implementation of deep learning and control algorithms with C++ language
- Network and communication