摘要
针对热连轧活套高度和张力系统的双输入双输出强耦合特性,提出了基于数据驱动的预测控制策略.利用离线和在线输入输出数据以及预测信息设计控制器,基于预测控制原理,在一个周期内,顺序更新活套高度和张力系统的控制律.该方法将一个复杂的系统分解成多个相关联的子系统设计控制器,降低系统的维数,减小的计算量,提高控制精度.仿真结果表明该方法具有较好的解耦效果和控制性能.
A decoupling control strategy based on data-driven predictive control theory is proposed for looper system of hot strip mills with strong coupling characteristics of double inputs and outputs . Based on subspace identification method, the controller is designed directlys using online and offline data, and the process of looper system model identification is avoided with improved ability of rejecting disturbance. Simulation results show that the control strategy is effective and has a better control performance.
出处
《大连交通大学学报》
CAS
2013年第4期69-72,共4页
Journal of Dalian Jiaotong University
基金
山东省自然科学基金资助项目(ZR2012FL16)
关键词
数据驱动
预测控制
活套系统
data driven
predictive control
looper system
