KDSource, a tool for the generation of Monte Carlo particle sources using kernel density estimation

Monte Carlo radiation transport simulations have clearly contributed to improve the design of nuclear systems. When performing in-beam or shielding a complexity arises due fact that particles must be tracked regions far from original source behind shielding, often lacking sufficient statistics. Different possibilities overcome this problem such as using particle lists generating synthetic sources already been reported. In work we present new approach by adaptive multivariate kernel density estimator (KDE) method. This concept was implemented in KDSource, general tool for modelling, optimizing and sampling KDE sources, which provides convenient user interface. The basic properties method were studied an analytical with known distribution. Furthermore, used two modelled neutron beams, showed good agreement experimental results.