Research On Deuteron Separation Energy And ~6Li(n,t)~4He Reaction Based On Bayesian Neural Network Approach

~6Li is a nuclide which has very unique structural properties in the 1p shell light core,and is the main raw material for tritium production.Therefor it has important practical value in nuclear engineering.Although the tritium production reaction ~6Li(n,t)~4He has made a significant breakthrough in experiments,its theoretical research is still progressing relatively slowly,which cannot meet the actual needs.So far,there is no set of theoretical models that can simultaneously and self-consistently describe the angular distribution and reaction cross the section of tritium production.However,The Bayesian Neural Network(BNN)approach is favored by many fields(including nuclear physics)by virtue of its high accuracy,by giving deviations from theoretical predictions,and by broad application prospects.The main research content of this paper is to try to establish a set of theoretical system and calculation tools for the application of BNN method to nuclear data evaluation.First,the deuteron separation energy is used as the research object to verify the reliability of the application of the BNN method in the field of nuclear quality research.The binding energy of nuclei is one of the basic properties of nuclei.The current research on binding energy mainly focuses on single neutron separation energy,single proton separation energy,double neutron separation energy,double proton separation energy and alpha particle separation energy.However,the research on the separation energy of deuterons has not yet been involved.In the study,it was found that the root mean square deviations of the three nuclear quality models processed by the method of BNN were reduced to varying degrees.In addition,we added physical quantities related to the pair effect and the shell effect into the input,which further improved the theoretical accuracy of the deuteron separation energy of the three nuclear mass models.The above results indicate that if more physical features are included in the BNN method,theoretical predictions with better accuracy of experimental data would be achieved.Based on the successful application of the BNN method to the separation energy of deuterons,the differential cross section of the output tritium of the ~6Li(n,t)~4He reaction is taken as the research object.Based on the latest measured differential Cross-section data issued by the Peking University Zhang Guohui research group in 2020(the incident neutron energy range is between 1 e V～3 Me V),we can processed it to reduce the order of magnitude difference in the complete set of data.The processed results are used as the learning set for learning,and the learning set and the verification set are in good agreement with the experimental data.The calculation method of the differential cross section of the ~6Li(n,t)~4He reaction and the deviation range of the theoretical value under any incident energy and exit angle will be given further.On this basis,the angle integral of the differential cross section is carried out,and the theoretical value of the reaction cross section of each incident energy point along with its theoretical deviation are self-consistently given,and it is in good agreement with the experimental measurement values??of different research groups in the world.Compared with the evaluation data of internationally famous databases such as ENDF/B-VIII.0 and JEFF3.3,the results of this paper are more self-consistent and can give the deviation range of theoretical prediction values,which is what the current nuclear data field cannot do,whether in theoretical methods or in evaluation methods.In this paper,the BNN method is used to study the deuteron separation energy and the ~6Li(n,t)~4He reaction.The results show that the BNN method has great development potential and application prospects in the field of nuclear structure and nuclear data research.