Zbigniew Moroz1, Jacek Boużyk2, Mieczysław Sowinski2, Andrzej G. Chmielewski3
1Spectroscopy and Nuclear Technology Division, The Andrzej Soltan Institute for Nuclear Studies, 05-400 Otwock/Swierk, Poland, and Bialystok Technical University, 45A Wiejska Str., 15-351 Bialystok, Poland,
2Spectroscopy and Nuclear Technology Division, The Andrzej Soltan Institute for Nuclear Studies, 05-400 Otwock/Swierk, Poland,
3Department of Nuclear Methods in Process Engineering, Institute of Nuclear Chemistry and Technology, 16 Dorodna Str., 03-195 Warsaw, Poland
The use of the computer genetic algorithm (GA) for driving a controller device for the industrial flue gas purification systems emploing the electron beam irradiation, has been studied. As the mathematical model of the installation the properly trained artificial neural net (ANN) was used. Various cost functions and optimising strategies of the genetic code were tested. These computer simulations proved, that ANN + GA controller can be sufficiently precise and fast to be applied in real installations.