| 摘要: | 摘要： The Rosenblueth point estimate method is one of the probabilistic analyses in estimating failure probability of a system, such as a slope. The essence of the approach is to use two point estimates, mean value±standard deviation, to present a variable in safety evaluation. The simple and straightforward framework leads to its wide application, but as a system governed by n variables (n is large), mass computations (2n repetitions in calculation) are required during the analysis. This prevents the possibility of hand computation using the approach, and a proper computing tool is needed under this situation. In this study, a Microsoft Excel-based program, RosenPoint, was developed for the Rosenblueth approach, and the program developments, descriptions and modifications are given in detail. The program is successfully demonstrated by computing the failure probability of an infinite slope under earthquake condition with a deterministic factor of safety (FOS) equal to 1.77. As the critical FOS is equal to 1.4, the slope that is considered stable by a conventional analysis is found associated with a substantial failure probability around 20%. Since the current version of RosenPoint is designed for estimating slope failure probability, the program needs modification as it is used for other tasks. Owing to the separated programming structure in RosenPoint, the subroutine governing FOS algorithms only needs to be replaced or recompiled as modification is needed. In addition, the capacity of the current RosenPoint is limited to 19 variables due to the dimension constraint of Excel spreadsheets (=220 rows). However, the capacity can be easily improved with sacrificing output completeness. This program modification is also described in this paper. ► An Excel-based program for the Rosenblueth point estimate method was developed. ► The program descriptions and modifications are included in this paper. ► The program was successfully used in evaluating slope failure probability. ► A stable slope through conventional analysis is found with 20% failure probability.
出版者： Elsevier Ltd
出版日期： 2012-11-01
出處： Computers & geosciences, 2012-11, Vol.48, p.239-243
資源來源： Elsevier ScienceDirect Journals Complete
版權： 2012 Elsevier Ltd
識別號： ISSN: 0098-3004
識別號： EISSN: 1873-7803
識別號： DOI: 10.1016/j.cageo.2012.01.009 |
