Hammerstein-Wiener模型最小二乘向量机辨识及其应用  

Hammerstein-Wiener model identified by least-squares-support-vector machine and its application

作　　者：桂卫华[1] 宋海鹰[1] 阳春华[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《控制理论与应用》2008年第3期393-397,共5页Control Theory & Applications

基　　金：国家自然科学基金(60634020;60574030);国家重点基础研究发展规划资助项目(2002cb312200);博士点基金(20050533016).

摘　　要：借鉴最小二乘支持向量机求解的思路,文中提出了辨识多输入-单输出Hammerstein-Wiener模型的方法.引入共线性约束假设,将辨识问题转换为有约束的优化问题,从而辨识出Hammerstein-Wiener模型的参数.基于Hammerstein-Wiener模型,我们建立了一个多输入-单输出的滚动预测模型,对铜转炉造渣S2期吹炼所需总氧量进行了预测,其相对均方根误差为12.1%.仿真结果表明,该模型预测准确、具有较好的应用价值.The identification method for a multi-input single-output Hammerstein-Wiener model is proposed by using the solving method of the least-squares-support-vector machine. The identification problem is converted into a constrained optimization problem by assuming collinear constraints so that the parameters of Hammerstein-Wiener model can be identified. Based on the Hammerstein-Wiener model, a multi-input single-output receding-horizon prediction model is developed for predicting the total oxygen quantity required by a copper converter in slag making S2 stage. The relative root-mean- square error （RRMSE） is 12.1%. The simulation research shows that this model provides accurate prediction and is with desirable application value.

关 键 词：Hammerstein—Wiener模型 最小二乘向量机 系统辨识 智能建模 铜转炉吹炼预测 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]