Genetic based sensorless hybrid intelligent controller for strip loop formation control between inter-stands in hot...

Area of Science:

• Industrial Process Control
• Mechanical Engineering
• Artificial Intelligence in Manufacturing

Background:

• Safe industrial operations are critical, especially in steel rolling mills where strip sag poses significant risks.
• Excessive strip sag can cause production scrap and damage to expensive machinery.
• Conventional controllers lack adaptability to real-world factors like temperature fluctuations and equipment wear.

Purpose of the Study:

• To develop and evaluate a novel controller for mitigating excessive strip sag in steel rolling mills.
• To improve the safety and efficiency of the hot strip rolling process.
• To address limitations of conventional controllers by incorporating adaptive elements.

Main Methods:

• A Genetic-neuro-fuzzy hybrid controller was designed and implemented without requiring a tension sensor.
• The proposed controller's performance was evaluated against fuzzy logic, neuro-fuzzy, and conventional controllers.
• Plant data was utilized for comparative analysis and simulation.

Main Results:

• The Genetic-neuro-fuzzy hybrid controller demonstrated superior performance in reducing excessive strip sag.
• The proposed controller effectively optimized sag control compared to existing methods.
• Simulation results validated the enhanced capabilities of the novel approach.

Conclusions:

• The developed Genetic-neuro-fuzzy hybrid controller offers a significant improvement for managing strip sag in steel rolling.
• This advanced control strategy enhances operational safety and reduces material waste.
• The controller's effectiveness, even without a tension sensor, highlights its practical applicability.