• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2015, Vol. 51 ›› Issue (4): 15-21.doi: 10.3901/JME.2015.04.015

• 仪器科学与技术 • 上一篇    下一篇

基于KICA子空间虚假邻点判别的软传感器变量选择方法

苏盈盈1, 2 李太福1 易 军1 胡文金1 廖志强3 徐 敏1   

  1. 1. 重庆科技学院电气与信息工程学院 重庆 401331
    2. 重庆大学自动化学院 重庆 400040
    3. 西安石油大学电子工程学院 西安 710065
  • 出版日期:2015-02-20 发布日期:2015-02-20
  • 基金资助:
    国家自然科学基金(61174015, 51376204, 51375520)、国家十二五科技支撑(2012BAK03B05)、重庆市自然科学基金(cstc2012jjA40026)和重庆科技学院校内重点基金(CK2011Z01)资助项目。

Variable Selection for Nonlinear Soft Sensor Based on False Nearest Neighbors in KICA Space

SU Yingying1, 2 LI Taifu1 YI Jun1 HU Wenjin1 LIAO Zhiqiang3 XU Min1   

  1. 1. School of Electrical and Information Engineering, Chongqing University of Science and Technology, Chongqing 401331
    2. College of Automation, Chongqing University, Chongqing 400040
    3. School of Electronic Engineering, Xi’an Shiyou University, Xi’an 710065
  • Online:2015-02-20 Published:2015-02-20

摘要: 针对软传感器建模过程中,高维冗余的非线性辅助变量造成的维度灾难问题,提出一种结合核独立成分分析法(Kernel independent component analysis, KICA)与虚假最近邻点法(False nearest neighbors, FNN)的非线性辅助变量选择方法。主要利用核函数将原始非线性数据映射到线性子空间,并采用独立成分分析消除因子之间的多重共线性,再运用虚假最近邻点法,计算原始数据在KICA子空间中投影的距离,依次判断各辅助变量对主导变量的解释能力,由此进行非线性变量选择。以某企业氢氰酸(Hydrocyanic acid, HCN)生产工艺过程中的转化率为软传感器预测目标,仿真结果表明该方法可有效降低辅助变量的维数、同时提高模型的预测精度。

关键词: 变量选择, 非线性系统, 核独立成分分析, 软传感器, 虚假最近邻点

Abstract: Selection of secondary variables is an effective way to reduce redundant information and to improve efficiency in nonlinear soft sensor. A novel method based on kernel independent component analysis (KICA) and false nearest neighbors method (FNN) is proposed on selecting the most suitable secondary process variables. The first step is to convert the non-linear operating variables into the linear space with kernel method. One the basis, they are projected into the independent ones with KICA transformations. In order to compare the different impacts on the operating variables, each original variable is eliminated orderly from original datasets with FNN in KICA subspace. In this way, it is possible to trace the important cause for the prediction. The result shows its validity with the verification in hydrocyanic acid (HCN) process industry.

Key words: false nearest neighbors, kernel independent component analysis, nonlinear system, soft sensor, variable selection

中图分类号: