基于灰色预测的MFAC在铝热连轧张力控制中的仿真研究

Simulation Study of Model-free Adaptive Controller Based on Grey Prediction in Tension Control of Aluminum Hot Tandem Rolling

针对重庆某铝加工厂双机架铝带热连轧机机架间张力波动问题,分析了影响张力波动的原因。为提高张力系统的控制精度,对传统PID控制器进行了算法改进,采用无模型自适应算法进行控制器设计,并加入灰色预测模型,通过超前预测来补偿系统延时及参数时变,形成新的控制律作用在被控对象上。仿真结果显示,基于灰色预测的无模型自适应控制在跟踪性、适应性、克服大时滞能力方面与传统的PID控制相比有很大的优越性。

Aimed at tension fluctuation between the standers of aluminum strip hot tandem rolling mill in an aluminum processing factory of Chongqing, analyses were made on the factors causing tension variation. In order to increase the control accuracy of tension system, algorithm was improved for the traditional PID controller, that is, the controller was designed based on model-free adaptive control algorithm, to which grey prediction model was introduced to compensate the time-delay of system and time-varying of parameter by prediction and form new control law acting on the controlled object. Simulation results show that the model-free adaptive control algorithm based on grey prediction has a lot of advantages in term of traceability, adaptability, ability to overcome large delay, compared with traditional PID control.