Optimization Strategy of Rolling Mill Hydraulic Roll Gap Control System Based on Improved Particle Swarm PID...

Area of Science:

• Materials Science and Engineering
• Control Systems Engineering
• Manufacturing Process Optimization

Background:

• Medium and heavy plates are critical materials for shipbuilding, defense, infrastructure, and construction.
• Traditional thick plate mill control systems lack the precision required for high-quality plate production, leading to thickness errors.
• Existing control technologies struggle to meet the high-precision roll gap control demands of thick plate mills.

Purpose of the Study:

• To develop and optimize an automatic control system for thick plate mill thickness.
• To enhance the precision and efficiency of plate thickness control on a 5500 mm thick plate production line.
• To compare the effectiveness of different PID controller optimization algorithms.

Main Methods:

• Modeling of the rolling mill plate thickness automatic control system for a 5500 mm production line.
• Optimization of PID controller parameters using Ziegler-Nichol (Z-N), Particle Swarm Optimization (PSO), and Linear Weight Particle Swarm Optimization (LWPSO) algorithms.
• Online semi-physical simulation using OPC UA communication technology, integrating Siemens S7-1500 PLC and MATLAB R2018b.

Main Results:

• The LWPSO algorithm demonstrated a 14.26% reduction in overshoot compared to Z-N and 10.18% compared to PSO.
• LWPSO advanced peak time by 0.31 s (vs. Z-N) and 0.05 s (vs. PSO), and reduced stabilization time by 3.71 s (vs. Z-N) and 4.31 s (vs. PSO).
• The optimized system shows improved control accuracy, speed, and anti-interference capabilities.

Conclusions:

• Linear Weight Particle Swarm Optimization (LWPSO) provides superior PID parameter tuning for thick plate mill thickness control.
• The developed control system effectively enhances product quality and production efficiency.
• The study offers valuable engineering insights and practical applications for the steel rolling industry.