Hot Strip Finishing Train Temperature and Geometry Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current finishing train systems in hot strip mills face challenges in real-time control of strip thickness and geometry due to the complexity of processing the entire system status, leading to inefficiencies in temperature and geometry tracking and correction.
Innovation Solution
A method that records initial temperature and geometry of hot strip points entering the finishing train, continuously updates actual temperatures and geometries in real-time using a model that accounts for temperature and geometry influences, and applies correction factors to adjust the strip points' geometries during passage, enabling continuous tracking and error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire system status of the finishing train is processed in real-time using complex control algorithms, then the accuracy of temperature and geometry tracking is improved, but the computational complexity and processing time exceed real-time requirements
Solution Approach 1:
The strip is divided into discrete strip points that are tracked individually through the finishing train. Each strip point maintains its own temperature and geometry data, allowing distributed computation rather than processing the entire system status as a single complex problem. This segmentation enables real-time processing by breaking down the computational burden into manageable per-point calculations.
Solution Approach 2:
Initial temperature and geometry values are recorded for each strip point as it enters the finishing train, and these values are assigned as actual values before processing begins. This preliminary assignment establishes a baseline that propagates through the system, allowing subsequent real-time updates to build upon pre-computed foundations rather than calculating everything from scratch at each time step.
2Adaptability or versatility
If complex control algorithms process the entire system status in real-time, then holistic control strategies are enabled, but the processing speed becomes too slow for real-time control cycles
Solution Approach 1:
By segmenting the control problem into individual strip point trajectories, the system achieves holistic control through coordinated management of many simple elements rather than processing the entire system as one complex entity. Each strip point is controlled independently based on its own path and conditions, yet the collective effect provides comprehensive system control.
Solution Approach 2:
The system performs preliminary tracking of each strip point's path through the finishing train and pre-assigns initial conditions. This allows the real-time control algorithm to focus only on updating temperature and geometry based on measured influences along the pre-determined path, rather than recalculating the entire trajectory at each control cycle.
3Productivity
If simplified models are used in basic automation for real-time processing, then processing speed is maintained, but only local statements are possible without global system awareness
Solution Approach 1:
The system uses simple local models for each strip point that only consider influences specific to that point's location and path. By segmenting the control problem this way, each local model remains computationally simple while the collection of all strip point models provides global system awareness through their coordinated evolution as they progress through the finishing train.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
At the latest when strip points (6) of a hot metal strip (5) enter a finishing train comprising a plurality of roll stands (F1...F6), a starting temperature (T1) and a starting geometry (G1) are captured for each strip point (6) entering and are assigned to the respective strip point (6) as the actual temperature (T2) and actual geometry (G2). The strip points (6) are tracked as they pass through the finishing train. The hot strip (5) is subjected to temperature and geometry influences (d?, dG) in the finishing train. By means of a model (9) and using the actual temperatures (T2), the actual geometry (G2) and the temperature and geometry influences (d?, dG), taking into account the tracking, new actual temperatures (T2) and new actual geometries (G2) of the captured strip points (6) are determined in real time and assigned to the detected strip points (6) such that they are available at any time during the passage of the strip points (6) through the finishing train. After the strip points (6) exit the finishing train, the final geometries (G3) thereof are captured. Using the respective captured final geometry (G3) and the actual geometry (G2), correction factors (Sk) are determined for modelling the roll stands (F1...F6) by means of the model (9). In addition, functional dependencies of the actual geometries (G2) on the correction factors (Sk) are determined by means of the model (9) such that the actual geometries (G2) of the strip points (6) which have already entered the finishing train can be corrected based on the correction factors (Sk) while passing through the finishing train.