Identification of trends in process measurements using the wavelet transform

被引:33
作者
Flehmig, F [1 ]
Von Watzdorf, R [1 ]
Marquardt, W [1 ]
机构
[1] RWTH Aachen, Lehrstuhl Prozesstech, D-52064 Aachen, Germany
关键词
D O I
10.1016/S0098-1354(98)00092-1
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
In conjunction with model based techniques for plant operation, advanced concepts are required to derive higher value information from plant measurements. Frequently, the qualitative behavior of the plant is of interest rather than sequences of measured data of high sampling rate. In this paper anew, wavelet based approach is suggested to identify and localize polynomial trends in noisy measurements. Due to a hierarchical search in the time-frequency plane, the method is highly computational efficient. It yields both the least-squares polynomials for the identified intervals and a quantitative measure for their goodness of fit. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.
引用
收藏
页码:S491 / S496
页数:6
相关论文
共 9 条
[1]  
[Anonymous], 1993, Ten Lectures of Wavelets
[2]   REPRESENTATION OF PROCESS TRENDS .3. MULTISCALE EXTRACTION OF TRENDS FROM PROCESS DATA [J].
BAKSHI, BR ;
STEPHANOPOULOS, G .
COMPUTERS & CHEMICAL ENGINEERING, 1994, 18 (04) :267-302
[3]  
Chui C.K., 1992, An introduction to wavelets, V1, DOI DOI 10.1109/99.388960
[4]   USING THE REFINEMENT EQUATION FOR EVALUATING INTEGRALS OF WAVELETS [J].
DAHMEN, W ;
MICCHELLI, CA .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 1993, 30 (02) :507-537
[5]   IDEAL SPATIAL ADAPTATION BY WAVELET SHRINKAGE [J].
DONOHO, DL ;
JOHNSTONE, IM .
BIOMETRIKA, 1994, 81 (03) :425-455
[6]  
DONOHO DL, 1995, J ROY STAT SOC B MET, V57, P301
[7]  
KUNOTH A, 1995, TEXAS A M PREPRINT
[8]   CHARACTERIZATION OF SIGNALS FROM MULTISCALE EDGES [J].
MALLAT, S ;
ZHONG, S .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1992, 14 (07) :710-732
[9]   A THEORY FOR MULTIRESOLUTION SIGNAL DECOMPOSITION - THE WAVELET REPRESENTATION [J].
MALLAT, SG .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1989, 11 (07) :674-693