Open Collections

Estimation and control of paper machine variables using wavelet packet analysis

University of British Columbia

The thesis discusses a new method in paper machine signal analysis using the wavelet packet transformation. It has been shown that this technique can be applied to three main areas in paper machine control such as the paper machine data analysis including estimation and compression, control of basis weight or moisture variations, and control performance monitoring. Compared to the traditional windowed Fourier transform, other filtering techniques, and wavelet transform, this method offers greater flexibility in analyzing signals with suitable time-frequency resolution. The Best Basis Algorithm has been introduced to obtain a best decomposition tree structure in decomposition of a given signal. The paper machine model for use in C D control has been derived from experimental bump test data. It has been clearly shown that the new method along with wavelet estimator produces results superior to standard analysis methods by reducing the error of estimation. In estimating more complex signals, however, the wavelet packet method is superior than the wavelet analysis since the best decomposition tree structure is chosen enabling the estimation to be more precise in the regions where frequency changes. When applied to industrial data, the wavelet packet estimation has provided more visual results by distinguishing the bump locations more evidently. Moreover, it has been shown that wavelet packet compression is more efficient than the wavelet case since it has achieved more compression ratio by yielding the combination of more zero coefficients and lesser mean square error. The thesis also presents the control scheme of the paper machine as well as an efficient technique to monitor the control performance on-line using both wavelets and wavelet packets. It has been shown that the calculation of performance index using the wavelet packet method is more precise since it offers finer frequency resolution.

Thesis/Dissertation

application/pdf

2009-04-30

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/7769

Electrical and Computer Engineering

University of British Columbia

UBCV

DSpace