|Table of Contents|

Soft-sensing Method Based on Lazy Learning Algorithm

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
2007年06期
Page:
679-683
Research Field:
Publishing date:
2007-12-30

Info

Title:
Soft-sensing Method Based on Lazy Learning Algorithm
Author(s):
WANG Qi-hong1PAN Tian-hong2ZOU Yun3
1.Department of Automation Control Engineering,Changzhou College of Information Technology,Changzhou 213164,China;2.School of Electrical and Information Engineering,Jiangsu University,Zhenjiang 212013,China;3.School of Automation,NUST,Nanjing 210094,China
Keywords:
lazy learn ing k-vector nearest ne ighbors sof-t sensing ester rate
PACS:
TP273
DOI:
-
Abstract:
In v iew o f lots of unmeasured variables in industrial process, a soft sensingm ethod using lazy learn ing is presented. The k-vector nearest ne ighbor is used to generate a ne ighbor o f current reg im e in order to enhance the predictive capability o f the orig ina l algorithm. The comp lex ity of the a-l gorithm is decreased by recursive least squares and the opt imal so lution is addressed by PRESS. Using th ism ethod tomode l ester rate o f a local chem ica l plan,t this paper obtains the max imum inaccuracy o f 0. 874 2% . The simu lation results show that the perfect genera lizat ion performance can meet high-precisionm easuring requ irements and the method is very understandable and easy to be implemented.

References:

[ 1] 刘兆磊, 徐振来, 张光义, 等. 基于非均匀采样间隔的数据关联算法研究[ J]. 南京理工大学学报, 2007, 31( 3): 327- 331.
[ 2] 仝卫国, 杨耀权, 金秀章. 神经网络的气体流量软测量模型研究[ J]. 中国电机工程学报, 2006, 26( 1): 66- 69.
[ 3] YanW W, ShaoH H, W angX F. So ft sensing mode-l ing based on suppo rt vector m ach ine and Bayes ian m ode l se lection [ J]. Compu ters and Chem ical Eng-i neering, 2004, 28( 10) : 1 489- 1 498.
[ 4] 颜学峰, 余娟, 钱锋. 基于自适应偏最小二乘回归的初顶石脑油干点软测量[ J]. 化工学报, 2005, 56( 8): 1 511- 1 515.
[ 5] 李凌. 软测量过程中的若干建模方法研究[ J] . 沈阳化工学院学报, 2006, 20( 2) : 118- 121.
[ 6] Bontem pi G, B irattariM, Bersin iH. Lazy lea rn ing fo r loca lm ode ling and con tro l design [ J] . Int J Contro ,l 1999, 72( 1 /2): 643- 658.
[ 7] Zheng Q, K im uraH. Just- in-tim e m ode ling fo r function prediction and its app lication [ J] . A sian Journa l o f Contro,l 2001, 3( 1): 35- 44.
[ 8] Zhang J, Y im Y-S, Y ang J. Inte lligent se lec tion o f instances for prediction functions in Lazy lea rning a lgorithm s [ J] . Artificial Intelligence Rev iew, 1997, 11( 7 /8): 175- 191.
[ 9] 潘天红, 李少远. 基于即时学习的非线性系统优化控制[ J] . 控制与决策, 2007, 22( 1): 24- 29.
[ 10] 张宇, 李柠, 黄道. 基于多神经网络模型的酯化反应软测量[ J] . 华东理工大学学报, 2005, 13( 2): 208- 211.
[ 11] 张跃军, 李喜敏, 曲文超. 醇醚型磺基琥珀酸酯盐系列的合成、结构与性能[ J]. 南京理工大学学报, 2001, 25( 6): 617- 621.

Memo

Memo:
-
Last Update: 2007-12-30