The Optical potential for neutron-nucleus scattering derived by Bayesian optimization

ベイズ最適化により導出された中性子-核子散乱光学ポテンシャル

渡辺 証斗*; 湊 太志   ; 木村 真明*; 岩本 信之  

Watanabe, Shoto*; Minato, Futoshi; Kimura, Masaaki*; Iwamoto, Nobuyuki

We are working on a combination of nuclear reaction calculation code CCONE and machine learning libraries to generate nuclear data and improve their accuracy. The angular distributions of elastic and inelastic scatterings to the first excited state on Fe at several incident energies were calculated using CCONE, and the optical potential parameters were optimized to reproduce the experimental data by Bayesian optimization. The optimized parameters were the depth of the real volume and imaginary surface parts of the potential, their energy dependence, radius, and diffuseness. Using the obtained optical potential, we estimated the angular distributions at energies different from those used on the optimization, and found that the results were in good agreement with the experiment data. In this presentation, we will introduce these calculation results and future prospects.