Artificial neural networks technology for neutron spectrometry and dosimetry

Abstract

Artificial Neural Network Technology has been applied to unfold neutron spectra and to calculate 13 dosimetric quantities using seven count rates from a Bonner Sphere Spectrometer with a (6)LiI(Eu). Two different networks, one for spectrometry and another for dosimetry, were designed. To train and test both networks, 177 neutron spectra from the IAEA compilation were utilised. Spectra were re-binned into 31 energy groups, and the dosimetric quantities were calculated using the MCNP code and the fluence-to-dose conversion coefficients from ICRP 74. Neutron spectra and UTA4 response matrix were used to calculate the expected count rates in the Bonner spectrometer. Spectra and H(10) of (239)PuBe and (241)AmBe were experimentally obtained and compared with those determined with the artificial neural networks.

Robust Design of Artificial Neural Networks Methodology in Neutron Spectrometry

“…[27] Researchers have using ANNs to unfold neutron spectra from BSS. [48] Figure 11, shows the classical approach of neutron spectrometry by means ANN technology starting from rate counts measured with BSS.…”

Evolutionary Artificial Neural Networks in Neutron Spectrometry

“…However, these methods still present the serious drawback of requiring a very expert user for their operation and the necessity to provide an initial guess spectrum for the deconvolution of the spectrum. To overcome these drawbacks, alternative approaches have been studied and proposed, to make an efficient neutron dosimetry, and several unfolding procedures combined with various types of experimental methods have been reported such as Genetic Algorithms (GA) (Freeman et al, 1999;Mukherjee, 2004), and Artificial Neural Networks (ANN) (Vega-Carrillo et al, 2007b;2006;2010). Many of the previous studies in neutron spectrometry and dosimetry by using the ANN approach have found serious drawbacks in the ANN design process itself, mainly in the proper determination of the structural and learning parameters of the networks being designed .…”

“…For the anterior, the nuclear research community needs approaches that implement ANN models faster than what is currently available. In consequence, more research has been suggested in order to overcome these drawbacks (Bedogni et al, 2007;Vega-Carrillo et al, 2007b;2006;2010). At present, one promising technique to design the structural and learning parameters of ANN is by introducing adaptation of network training using EA.…”

Determination and mapping the spatial distribution of radioactivity of natural spring water in the Eastern Black Sea Region by using artificial neural network method

“…Such structures are processing structures that calculate the weight of connections on the system. In recent years, artificial neural network technique is used in many different areas, and quite successful results are obtained even in complex problems (Amanatiadis et al 2014;Bagheri et al 2014;Das et al 2015;Falamaki 2013;Farah et al 2011;Negarestani et al 2002;Otağ et al 2015;Raza et al 2012;Vega-Carrillo et al 2007;Zjavka 2014). The purpose of ANN method is making proper predictions for samples never seen before by using known samples collected before or previously measured results of the region.…”