ABSTRACT

A cross-section adjustment method based on Bayesian theory was derived for adjusting the specific cross-section set. The present method gives the ‘a posteriori’ distribution of the specific cross-section and does not require the adjustment of the other cross-sections. In the derivation, the experimental error, calculation error, and cross-section error are assumed to distribute normally and to be independent of each other. The adjusted specific cross-sections and the covariance between them are the same as those obtained by the conventional cross-section adjustment method. In addition, the adjustment of covariance between the adjusted cross-section and the other cross-section is not necessarily required in the derivation. For introducing these features of the present method, the specific cross-sections were adjusted by using the criticality of the light-water moderated uranium benchmark. The result shows that the uncertainty of the multiplication factor is reduced by the present method, and the reduction is drastically smaller than that by the conventional method.

摘要

推导出一种基于贝叶斯理论的截面调整方法，用于调整特定截面集。本方法给出了特定横截面的“后验”分布，不需要调整其他横截面。在推导中，假设实验误差、计算误差和横截面误差呈正态分布且相互独立。调整后的特定截面和它们之间的协方差与常规截面调整方法获得的相同。另外，在推导中不一定需要对调整后的横截面与其他横截面之间的协方差进行调整。为了介绍本方法的这些特点，使用轻水慢化铀基准的临界性调整了特定横截面。结果表明，本方法降低了乘法因子的不确定性，且降低幅度比传统方法小得多。