8 January 2008 Application of improved BP neural network in infrared spectroscopy
Author Affiliations +
Abstract
Artificial neural network(ANN), as a computing model possessing high-nonlinear mapping ability, has a wide application future in engineering. In this paper, combined with some practical problems in infrared spectroscopy, the method of quantitative analysis of the infrared spectrum using the improved BP-NN is discussed. The non-linear relationships between the component concentrations and the peak strengths are obtained. The method was tested and the resulted showed that the system performance was better than with several other commonly used conventional methods. Finally the problems which should be paid attention to are indicated and some helpful advice is given.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ming-zheng Gao, Ming-zheng Gao, Huo-ming Zhang, Huo-ming Zhang, Shang-zhong Jin, Shang-zhong Jin, } "Application of improved BP neural network in infrared spectroscopy", Proc. SPIE 6835, Infrared Materials, Devices, and Applications, 68352H (8 January 2008); doi: 10.1117/12.758211; https://doi.org/10.1117/12.758211
PROCEEDINGS
11 PAGES


SHARE
Back to Top