基于多支持向量机模型和优化控制器的板形、板厚控制  

Strip shape and gauge control based on multiple SVM model and optimization controller

作　　者：陈治明[1] 曹建忠[1] 

机构地区：[1]惠州学院电子科学系,广东惠州516007

出　　处：《控制与决策》2012年第4期525-530,共6页Control and Decision

基　　金：国家自然科学基金项目(60774032)

摘　　要：针对带钢热连轧过程中互相耦合的板形、板厚控制问题,提出一种综合控制策略.首先,在输入空间划分的基础上建立包含多个子模型的多支持向量机模型,并通过主元分析方法实现模型输出的综合;然后,利用建立起来的模型设计优化控制器,对板形、板厚进行综合控制.计算机仿真和现场实验结果均表明了所提出的基于多支持向量机模型的综合控制策略能同时有效地减小板形、板厚偏差.Aiming at the coupled shape and gauge control problem in hot strip mills, a complex control strategy is proposed. Based on the division of the input space, a multiple support vector machine model with several sub-models is established, and the principal component regression method is used for the output synthesis of the sub-models. Based on the established model, an optimization controller is designed for the complex control of strip shape and gauge. Simulation and field experiments show that the proposed complex control method based on the multiple support vector machine model can effectively reduce the ~au^e and shape control error at the same time.

关 键 词：板形 板厚控制 支持向量机 减法聚类 主元分析 优化 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]