International Symposium on Symbiotic Nuclear Power (ISSNP2024)

The purpose of implementing matrix correction is to compensate for the fissile nuclide mass measurement deviation caused by the matrix effect within a limited time frame. In previous work, we introduced a novel matrix correction method using the neutron flux monitor group (NFMG) response and applied it to an active neutron interrogation (ANI) system based on the delayed neutron. This method can be used to identify the known matrix by an improved simulated annealing (SA) algorithm and to correct the unknown matrix by fitting the equation. In this work, we further extend this method to ANI systems based on differential die-away (DDA) methods. To validate again the applicability and accuracy of the correction method, a DDA equipment model and various matrix compositions were developed using the Geant4 toolkit. The results demonstrate that for U3O8 materials with varying enrichment and distribution, the improved SA algorithm can still efficiently identify the various matrix materials in the reference database, and the identification rate is not less than 94.6%. Moreover, the average relative deviation of the corrected 235U mass is no more than 34.0% for the unknown matrix. This work effectively demonstrates the universality of the matrix correction method based on NFMG response for different ANI systems.

Nuclear safeguard;Differential die-away;Monte Carlo simulation;Matrix effect correction;Machine learning