engleski

Draft Version of the Computational Module for Simplified Radiological Analyses

In the framework of the DOCPAGANSA (“Development of Code Package for Advanced Gamma and Neutron Shielding Analyses”) research project the computational module for simplified shielding analyses has been developed. The computational module is based on the point kernel method which is used for gamma, as well as neutron shielding calculations. The gamma and neutron buildup factors are determined by a novel method based on Support Vector Regression method which is a one of the machine learning techniques. In this paper we report on the main characteristics of the developed preliminary SVR models for gamma and neutron buildup factors. We also present the preliminary results of the testing of the models’ beta versions on the geometrically simple problem setup with gamma and neutron source.

Shielding, neutron, gamma, point kernel, buildup factors, SVR method, machine learning

nije evidentirano