Z. Moroz1, M. Sowiński1, J. Boużyk1, A. G. Chmielewski2
1 The Andrzej Soltan Institute for Nuclear Studies, 05-400 Otwock/Swierk, Poland,
2 Institute of Nuclear Chemistry and Technology, 16 Dorodna Str., 03-195 Warsaw, Poland
The artificial neural net (ANN) technique has been studied as a tool for model calculations and control of the industrial exhaust gas purification systems employing the electron beam irradiation. As a simulator of multiparameter events from the installation, a simple analytical model was applied giving a reasonably good agreement with experimental data available. Various learning strategies of the ANN have been tested. Advantage of the ANN technique over other methods is discussed.