An application of wavelet transform for decomposition of millimeter-wave spectroscopic signals [electronic resource].
- Washington, D.C. : United States. Dept. of Energy, 1994.
Oak Ridge, Tenn. : Distributed by the Office of Scientific and Technical Information, U.S. Dept. of Energy.
- Physical Description:
- 9 pages : digital, PDF file
- Additional Creators:
- Argonne National Laboratory, United States. Department of Energy, and United States. Department of Energy. Office of Scientific and Technical Information
- Restrictions on Access:
- Free-to-read Unrestricted online access
- Millimeter-wave technique, based on rotational energy transitions of molecules, holds promise for remote monitoring of environmentally hazardous effluents from processes. Argonne National Laboratory is developing a millimeter-wave sensor based on active swept-frequency radar technique in the frequency range of 220-320 GHz. Because the line widths of millimeter-wave spectra of molecules at atmospheric pressure are broad (∼ 4 GHz half-width at half height), the composite spectrum of multicomponent mixtures of chemicals is generally complex and overlapping. This paper presents an application of discrete wavelet transform for efficient representation and decomposition of millimeter-wave spectral data. A two-layer back propagation neural network is trained using multifrequency wavelet coefficients of the signals as input features and the known composition of different chemicals in the mixture as target output vectors. After training, composition of an unknown mixture of the base chemicals is determined using the wavelet representation of its absorption spectra. Simulated and experimental spectral data were used to test the wavelet transform technique. Accurate values of individual chemical compositions resulted for noise-free laboratory data. In addition, the technique showed more robustness than conventional multivariate techniques under noisy conditions.
- Report Numbers:
- E 1.99:anl/et/cp--83924
E 1.99: conf-9405229--1
- Other Subject(s):
- Published through SciTech Connect.
1995 IEEE international conference on acoustics, speech and signal processing,Detroit, MI (United States),8-12 May 1994.
Gopalsami, N.; Raptis, A.C.; Bakhtiari, S.; Gopalan, K.
- Funding Information:
View MARC record | catkey: 14140159