Journal of Systems Engineering and Electronics ›› 2009, Vol. 31 ›› Issue (8): 1935-1939.

• 制导、导航与控制 • 上一篇    下一篇

基于小波基函数和Hammerstein模型的预测函数控制

刘春波1, 王鲜芳1,2, 章瑶1, 潘丰1   

  1. 1. 江南大学通信与控制工程学院, 江苏, 无锡, 214122;
    2. 河南科技学院信息工程学院, 河南, 新乡, 453003
  • 收稿日期:2008-06-03 修回日期:2009-04-15 出版日期:2009-08-20 发布日期:2010-01-03
  • 作者简介:刘春波(1980- ),男,博士研究生,主要研究方向为数据挖掘、人工智能.E-mail:wuyuchunbo@yahoo.com.cn
  • 基金资助:
    国家高技术研究发展计划(863计划)(2006AA020301)资助课题

Predictive functional control based on wavelet function and Hammerstein model

LIU Chun-bo1, WANG Xian-fang1,2, ZHANG Yao1, PAN Feng1   

  1. 1. School of Communications and Control Engineering, Jiangnan Univ., Wuxi 214122, China;
    2. School of Information Engineering, Henan Inst. of Science and Technology, Xinxiang 453003, China
  • Received:2008-06-03 Revised:2009-04-15 Online:2009-08-20 Published:2010-01-03

摘要: 针对实际对象的多变性和传统的预测函数控制只能对预测时域内有限的几个拟合点进行拟合,而未考虑整个预测时域整体逼近误差性能优化的问题,提出一种基于小波基函数和Hammerstein模型的预测函数控制方法.内部模型参数通过不断辨识,进行自适应校正.利用小波的紧支局部性和多尺度分析特性,既保证了整体误差性能的优化,又突出了重要拟合点的逼近要求,并实现了优化变量的集结.理论分析和仿真应用表明,该方法有更好的快速性和抗模型失配性能.

Abstract: Aiming at the problems that actual objects are changeful and traditional predictive functional control can only fit limited points in the time domain of prediction,being unable to optimize the target function in the whole time domain of prediction,a predictive functional control method based on wavelet function and Hammerstein model is proposed.The inner model parameters are adjusted automatically depending on repeated identifications.The wavelet properties of compact support and multi-scale analysis can ensure the target function of the whole time domain optimization,some important points fitting and the optimal parameters aggregation.The theoretical analysis and simulation show the approach has better rapidity and the ability of model's anti-mismatch.

中图分类号: