Multilevel Monte Carlo methods for computing failure probability of porous media flow systems

详细信息   

• 作者：F. Fagerlunda ; fritjof.fagerlund@geo.uu.se" class="auth_mail" title="E-mail the corresponding author ; F. Hellman ; b ; fredrik.hellman@it.uu.se" class="auth_mail" title="E-mail the corresponding author ; A. Må ; lqvistc ; axel@chalmers.se" class="auth_mail" title="E-mail the corresponding author ; A. Niemia ; auli.niemi@geo.uu.se" class="auth_mail" title="E-mail the corresponding author
• 关键词：CDF estimation ; Failure probability ; Porous media flow simulation ; Multilevel Monte Carlo ; Selective refinement
• 刊名：Advances in Water Resources
• 年：2016
• 期：Complete
• DOI：10.1016/j.advwatres.2016.06.007
• 来源：Elsevier
• 类型：期刊

摘要

•Multilevel Monte Carlo (MC) methods for estimating failure probability were assessed.•Computable error bounds for the quantity of interest allows for selective refinement.•Selective refinement combined with MC methods yields significant performance gains.