From “Old Antique” to Intelligent System: A Record of the Transformation of the S7-400 Rolling Mill Control System in the Metallurgical Industry

Do you remember the urgent call I received three years ago? “Engineer Li, the No. 3 rolling mill has stopped again, the old system just can’t hold up anymore!” At that time, I was debugging another project on-site, and upon receiving this call, I knew that the control system that had been in service for nearly 20 years had finally reached its end.

Project Background

A large cold rolling production line in Hubei was originally using a control system from the 1990s, which had an increasing failure rate and spare parts were almost obsolete. The most critical issue was that each failure meant losses of hundreds of thousands, even affecting downstream production.

In industrial sites, reliability is always the top priority. That old system could no longer meet the demands of modern production, and a transformation was imperative.

System Hardware Configuration

After multiple rounds of technical discussions, we ultimately chose the Siemens S7-400 as the core of the new system. The configuration is as follows:

• CPU: Using the 414-5H redundant CPU, this choice is definitely the king of cost-performance.
• Power Module: PS407 10A dual power supply redundancy configuration.
• Communication Processor: CP443-1 for host communication, CP443-5 Extended for Profibus-DP communication.
• I/O Configuration:
• Local I/O: SM431/SM432 analog modules.
• Remote I/O: Distributed in various control cabinets via ET200M stations.

To be honest, although this hardware configuration requires a significant investment, considering the high-speed response requirements of rolling mill control, this choice is pragmatic.

Key Points of System Design

What is the core of the rolling mill control system? That’s right, it’s real-time performance and reliability!

We adopted a three-layer control architecture:

1. 🔝 Upper Layer: MES system, responsible for production scheduling and process parameter distribution.
2. 🔄 Middle Layer: S7-400 PLC system, responsible for core control logic.
3. 🔌 Lower Layer: Field devices such as inverters, servo systems, and I/O devices.

The key to the control system lies in thickness control and tension control, as these two directly determine product quality. To achieve this, we designed a composite control strategy combining feedforward with PID, achieving a thickness precision control of ±0.01mm.

Program Implementation Details

I paid special attention to modular design in the program structure:

• OB1: Main loop program, responsible for overall scheduling.
• OB35: 100ms cycle interrupt for PID calculation.
• OB82/83/86: Diagnostic interrupts and hot restart interrupts to improve system stability.
• FB100~FB120: Various functional modules of the rolling mill, including thickness control, tension control, speed synchronization, etc.
• DB100~DB200: Relevant data blocks, storing process parameters and intermediate variables.

The core control algorithm combines classic PID with feedforward control. The thickness control uses the empirical formula accumulated by our team over the past decade:

Output = Kp*(Error) + Ki*∫Error dt + Kd*(d Error/dt) + Kf*Feedforward

In practical applications, it is far more complex, considering a series of mechanisms such as dead zone, filtering, and anti-integral saturation.

Debugging Experience Summary

This is the most grueling stage! We adopted a “four-step” strategy:

1. ✅ Unit Testing: Each functional module is validated individually.
2. ✅ Integration Testing: Simulating I/O signal tests without connecting to the drive.
3. ✅ No-load Operation: Testing mechanical actions without loading.
4. ✅ Load Operation: Gradually increasing speed and load until full load operation.

The biggest challenge came from the jitter problem in inverter communication. After repeated testing, we ultimately resolved this issue by optimizing Profibus network parameters and adding fiber optic isolators.

Project Benefits

After the transformation, the reliability of the production line improved from the original 92% to 99.7%, and the average downtime decreased from 280 hours to less than 30 hours per year. The first-pass yield of product quality increased by 3.2 percentage points, saving the company approximately 8 million yuan annually.

Thought-provoking Question: If your rolling mill system experiences thickness fluctuations, what aspects should you first investigate? Think about it, then consider my suggestions: first check if the PID parameters are appropriate, then see if the tension control is stable, and finally, check the sensor accuracy.

In the field of industrial control, one must not only understand technology but also the processes. I hope my experience sharing is helpful to you. If you encounter issues in S7-400 applications, feel free to discuss. In the path of industrial automation, we are all fellow travelers growing together.

Like and Share

Let Money and Love Flow to You