p.186
p.191
p.195
p.203
p.209
p.213
p.220
p.224
p.232
DWT-PLS Regression on Near-Infrared Spectra for Moisture Determination of Coal
Abstract:
We studied moisture determination in bituminous coal and lignitic coal samples using near-infrared (NIR) spectra. This research was developed by applying partial least squares regression (PLS) and discrete wavelet transform (DWT). Firstly, the NIR spectra were pre-processed by DWT for fitting and compression. Then, the compressed data were used to build regression model with PLS for moisture determination in coal samples. Compression performance at different resolution scales was investigated. Using the compressed data, PLS can obtain more accurate result than using raw spectra. The number of principal component in PLS model was investigated too. The results show DWT-PLS can obtain satisfactory determination performance for moisture analysis in bituminous coal and lignitic coal.
Info:
Periodical:
Pages:
209-212
Citation:
Online since:
October 2013
Authors:
Price:
Сopyright:
© 2014 Trans Tech Publications Ltd. All Rights Reserved
Share:
Citation: