Wojcik W., Gotra Z., Duk M., Kotyra A., Przylucki S.W., Smolarz A.
In this paper we present optoelectronic measurement system for monitoring the industrial gas pollutants. The system consists of a source of light, an optical fiber as a data transmission link, a spectrometer, an optical detection system and a neural network unit for a real-time spectral data processing. We have paid the major attention to the neural network structure and its properties in gas recognition and gas concentration estimation.
Key words: neural network, power industry, spectroscopy, environmental pollution
doi 10.3116/16091833/2/1/46/2001
References
1. Par Ragnarson, 1994. "Optical Techniques for Measurements of Atmospheric
Trace Gases" Lund Raports on Atomic Physics LRAP-152, Lund Institute of
Technology, Department of Physics, April.
2. Keller Paul E , Richard T Kozues, Kan-gas Lars J, 1993. Procceding
of the 8-th Con-fe-ren-ce on Real-Time Computer Appli-ca-tions in Nuclear,
Particle and Plasma Physics.
3. Krolopp WJ, Wójcik W, Kotyra A, Przy?ucki
S, Smolarz A, 1998. Reports of the Lodz Polytechnick, Electrical 90: 5–38.
4. Kohonen T, 1995. Self-Organizing Maps, Spriger Series in Information
Sciences.
5. Xiang Ouyang, Philip L Varghese, 1989. Appl. Optics 38.