Hot Strip Mill Command Speed Prediction for Temperature Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for maintaining a consistent final rolling temperature in hot strip mills are inaccurate due to limitations in predicting and adjusting the command speed, leading to deviations in the energy content of the strip at the coiler, as they rely on predefined acceleration ramps and single-step predictions, which fail to account for variations in command speed direction and value.
Innovation Solution
A method where a control computer determines command variables for strip points based on actual and setpoint energy values, using a specific determining rule to adjust command speeds before the strip enters the production line, and continuously adjusts operation to match expected energy content values at the coiler, incorporating a data field for efficient computation and real-time adaptation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If predefined acceleration ramps are used to control command speed, then the control system is simple to operate, but the final rolling temperature accuracy deteriorates due to inability to account for command speed variations
Solution Approach 1:
The system performs preliminary prediction of command speed values for multiple future strip points before the strip enters the production line. The control computer calculates expected energy content and determines optimal command speed profiles in advance, allowing the system to proactively compensate for temperature variations rather than reacting to them after they occur.
Solution Approach 2:
The system transitions from static predefined acceleration ramps to dynamic command speed adjustment. The command speed for each strip point is individually optimized based on predicted energy content, actual measured values, and determining rules that adapt to real-time conditions, enabling the control system to dynamically respond to variations in strip properties and production conditions.
2Device complexity
If single-step prediction methods are used for command speed, then the computational complexity is low, but the energy content consistency at the coiler deteriorates due to failure to account for command speed direction and value variations
Solution Approach 1:
The prediction process is segmented into multiple discrete steps, with command speed predicted for multiple future strip points (not just the next single point). Each strip point has its own predicted command speed value based on its specific position and conditions, allowing the system to account for variations in command speed direction and magnitude throughout the strip's passage through the production line.
Solution Approach 2:
The system maintains continuous energy content control by consistently applying determining rules across multiple strip points. The control computer continuously compares expected energy content with actual values and adjusts command speeds in a continuous manner, ensuring stable energy content at the coiler throughout the rolling process rather than relying on intermittent corrections.
3Loss of time
If command speed is adjusted based on current strip point only, then the response time is fast, but the temperature control accuracy deteriorates due to lack of consideration for subsequent strip points
Solution Approach 1:
The control computer determines command speed values for multiple future strip points in advance, before those strip points actually enter the production line. This preliminary determination allows the system to prepare optimal control actions ahead of time, maintaining fast response while ensuring accurate temperature control for each subsequent strip point based on its specific conditions.
Data Source
AI summary
Before a first strip point is fed into a production line, an actual energy content at a location in front of the production line and a setpoint energy content at a location behind the production line are received for a first strip point, second strip point, and third strip point. The third strip point, followed by the first strip point, followed by the second strip point, are fed into the production line. A command variable for the first strip point and second strip point(s) is determined prior to feeding in the first strip point. Each command variable is determined based on (a) the actual value and the setpoint value of the strip point currently entering the production line, and (b) the actual value and the setpoint value of at least one strip point having already entered.