化工学报 ›› 2012, Vol. 63 ›› Issue (9): 2887-2891.DOI: 10.3969/j.issn.0438-1157.2012.09.034

• 过程系统工程 • 上一篇    下一篇

基于递推PLS核算法的软测量在线学习方法

邵伟明, 田学民, 王平   

  1. 中国石油大学(华东)信息与控制工程学院, 山东 青岛 266580
  • 收稿日期:2012-06-14 修回日期:2012-06-21 出版日期:2012-09-05 发布日期:2012-09-05
  • 通讯作者: 田学民
  • 作者简介:邵伟明(1986-),男,博士研究生。
  • 基金资助:

    国家自然科学基金项目(51104175);山东省自然科学基金项目(ZR2011FM014)。

Online learning soft sensor method based on recursive kernel algorithm for PLS

SHAO Weiming, TIAN Xuemin, WANG Ping   

  1. College of Information and Control Engineering, China University of Petroleum, Qingdao 266580, Shandong, China
  • Received:2012-06-14 Revised:2012-06-21 Online:2012-09-05 Published:2012-09-05
  • Supported by:

    supported by the National Natural Science Foundation of China(51104175)and the Natural Science Foundation of Shandong Province(ZR2011FM014).

摘要: 针对过程的动态时变特性,提出一种基于PLS核算法的软测量在线学习方法。该方法利用PLS核算法,通过递推学习具有代表性的新样本来改善模型的适应能力,较NIPALS算法具有更高的计算效率;并采用一种同时考虑输入和输出信息的相似度准则,有选择地删除一个或多个冗余样本,更有效地构建了训练样本集。工业聚丙烯熔融指数的软测量建模研究表明,本文提出的方法能够快速有效地跟踪牌号切换中熔融指数的变化。

关键词: 软测量, 在线学习, PLS核算法, 样本相似度

Abstract: An online learning soft sensor method based on PLS kernel algorithm is presented for industrial processes with time-varying characteristics.By recursively learning representative samples,this method,utilizing PLS kernel algorithm,could improve soft sensor model’s ability of adaptation,which is more computationally efficient than NIPALS.And deleting one or more redundant samples according to a similarity-based criteria that takes both input and output information into consideration may build more effective training sample set.The foregoing scheme is applied to build an industrial polypropylene unit’s melt index model.The results indicate that the proposed method can efficiently track the change of melt index during grades transition.

Key words: soft sensor, online learning, kernel algorithm for PLS, sample similarity

中图分类号: