Robustness Analysis Of AGC System In Tandem Cold Mill

The AGC (Automatic Gauge Control) system of tandem cold mill features multiple variable, strong coupling, non-linearity and time delay. It is among the most difficult technology problems in automatic system design of tandem cold mill. The industrial AGC system typically employs decoupling control (PID controller plus compensation). The variables required to control in production include rolling force, strip tension, roll speed, strip speed and flatness. There exist coupling relations among different control targets, e.g. gauge-flatness, speed-tension, speed-gauge, etc. By using PID decoupling control, the control system has a simple structure and is easy to evaluate. It also has a certain level of robust stability. With the increasing demand on product quality and yield requirement, it is urgent to analyze the AGC system robustness of tandem cold mill.In the paper, targeted at the most widely used five stand tandem cold mill, an introduction is given on basic rolling principles, rolling force model, foward slip model, tension model and interstand gauge delay model. In addition, the common units in AGC system are introduced to provide a basis for robustness analysis of AGC system.AGC systems in tandem cold mill can be described as low order systems with time delay. A software system targeted at industrial plants is designed to incorporate system identification for plant model and PID controller design based on the identified model. The system identification algorithms, controller design algorithms and GUI of the software, which is able to run independently, are implemented in MATLAB. New methods are put forward, with regard to model order identification criterion, model approximation of first order lag plus delay system, as well as saturation and non-linearity manipulation of PID controller design. A case study of mass flow AGC system is introduced, which verifies the identification algorithms, controller performance, and software. Satisfactory controlling effects could be attained by employing the designed software.After acquiring a system model with system identification methods, different PID controllers could be designed with different design algorithms. Even if the designed controllers prove to have outstanding control performance on given steady model, the performance could degenerate greatly in case of parameter disturbance. Robustness analysis on analog PID controller and digitial PID controller is elaborated and a design method of PID parameters is put forward to meet the requirements of robustness. The theory and design method have important practical significance on AGC control system design of five stand tandem cold mill.