Spectra and dose with ANN of Cf, Am--Be, and Pu--Be

• Published in: Journal of radioanalytical and nuclear chemistry Vol. 281; no. 3; pp. 615 - 618
• Main Authors: Vega-Carrillo, Hector Rene, Martinez-Blanco, M R, Hernandez-Davila, Victor Martin, Ortiz-Rodriguez, Jose Manuel
• Format: Journal Article
• Language: English
• Published: 01-09-2009
• Subjects: Artificial neural networks; Equivalence; Learning theory; Neural networks; Neutron sources; Neutron spectra; Nuclear chemistry; Spectra
• ISSN: 0236-5731; 1588-2780
• DOI: 10.1007/s10967-009-0057-z

Artificial neural networks have been applied to unfold the neutron spectra and to calculate the effective dose, the ambient equivalent dose, and the personal dose equivalent for Cf, Am--Be, and Pu--Be neutron sources. The count rates that these neutron sources produce in a Bonner Sphere Spectrometer with a LiI(Eu) were utilized as input in both artificial neural networks. Spectra and the ambient dose equivalent were also obtained with BUNKIUT code and the UTA4 response matrix. With both procedures spectra and ambient dose equivalent agrees in less than 10%. The Artificial neural network technology is an alternative procedure to unfold neutron spectra and to perform neutron dosimetry.