期刊文献+
共找到1篇文章
< 1 >
每页显示 20 50 100
History-by-History Variance in Monte Carlo Simulation of Radiation Interactions with Matter
1
作者 mary pik wai chin 《Applied Mathematics》 2017年第3期293-306,共14页
Radiation interaction with matter is by nature stochastic. Monte Carlo simulation makes possible, practical, economical and ethical, many experiments otherwise not. In-silico simulation of random samples of radiation ... Radiation interaction with matter is by nature stochastic. Monte Carlo simulation makes possible, practical, economical and ethical, many experiments otherwise not. In-silico simulation of random samples of radiation histories places into our hands data which may be treated and analysed in radically different ways. This work reports history-by-history computation of the variance in simulation output from FLUKA, a general-purpose Monte Carlo code. Variance computed history by history is compared with variance estimated from varying batch sizes. This work also addresses the issue of under-sampling where, against conventional expectations, the variance spikes as we sample more histories. The background to the spikes is traced by reconstructing single histories interaction by interaction—a novel level of details atypical of Monte Carlo simulations in the field, where statistical convergence of averaged quantities has been the focus. 展开更多
关键词 MONTE Carlo FLUKA Error STATISTICS Variance
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部