可逆冷带轧机速度张力多变量耦合系统的建模及分散控制  

Modeling and decentralized control for speed and tension multivariable coupling system of reversible cold strip mill

作　　者：刘乐[1] 方一鸣[1,2] 李建雄[1] 常茹[1] 

机构地区：[1]燕山大学电气工程学院工业计算机控制工程河北省重点实验室,河北秦皇岛066004 [2]国家冷轧板带装备及工艺工程技术研究中心,河北秦皇岛066004

出　　处：《控制理论与应用》2014年第1期42-48,共7页Control Theory & Applications

基　　金：国家自然科学基金资助项目(61074099);河北省高等学校创新团队领军人才培育计划(LJRC013);燕山大学博士基金资助项目(B705)

摘　　要：针对具有多变量、非线性、强耦合和不确定性的可逆冷带轧机速度张力系统,提出了一种基于扩张状态观测器(extended state observer,ESO)的全局积分滑模自适应反步分散控制方法.首先,采用机理建模方法,建立了相对完备的可逆冷带轧机速度张力多变量耦合系统的数学模型.其次,将各子系统的耦合项和不确定项看成外扰,通过构造的ESO对其进行动态观测,并分别引入所设计的全局积分滑模自适应反步控制器中进行补偿,速度张力系统实现了有效的动态解耦和协调控制.理论分析表明,所提出的控制方法能够保证滑模面的渐近稳定和闭环系统的渐近跟踪性能.最后,基于某1422mm可逆冷带轧机速度张力系统的实际数据进行仿真,结果验证了所提方法的有效性.For the speed and tension system of reversing cold strip mill with multivariable, nonlinearities, strong cou- pling and uncertainties, a global integral sliding mode adaptive backstepping decentralized control strategy based on ex- tended state observer （ESO） is proposed. Firstly, a relatively complete mathematical model for the speed and tension multivariable coupling system is built by using the mechanism modeling approach. Then, regard the coupling terms and uncertain term of each subsystem as external disturbance, for which we build ESO to conduct dynamic observation, and the observed value is introduced into the global integral sliding mode adaptive backstepping controller to carry on compensa- tion, then to achieve effective dynamic decoupling and coordination control for the speed and tension system. Theoretical analysis shows that the sliding mode surfaces are asymptotically stable, and the speed and tension closed-loop system has desirable asymptotic tracking performance. Finally, a simulation is carried out on the speed and tension system of a 1422 mm reversing cold strip mill by using the actual data. Results show the validity of the proposed control strategy.

关 键 词：可逆冷带轧机 速度张力系统 分散控制 动态解耦 全局积分滑模 自适应反步控制 扩张状态观测器 

分 类 号：TG334.9[金属学及工艺—金属压力加工] TP273[自动化与计算机技术—检测技术与自动化装置]