Analytical Results of Variance Reduction Characteristics of Biased Monte Carlo for Deep-Penetration Problems
- 1 March 1986
- journal article
- research article
- Published by Taylor & Francis in Nuclear Science and Engineering
- Vol. 92 (3) , 482-487
- https://doi.org/10.13182/nse86-a17535
Abstract
An analytical formulation is presented for calculating the mean and variance of transmission for a model deep-penetration problem. With this formulation, the variance reduction characteristics of two biased Monte Carlo schemes are studied. The first is the usual exponential biasing wherein it is shown that the optimal biasing parameter depends sensitively on the scattering properties of the shielding medium. The second is a scheme that couples exponential biasing to the scattering angle biasing proposed recently. It is demonstrated that the coupled scheme performs better than exponential biasing.Keywords
This publication has 8 references indexed in Scilit:
- Asymptotic survival probability of random walks on a semi-infinite linePramana, 1985
- Optimal choice of parameters for tracklength biassing in Monte Carlo deep-penetration problemsAnnals of Nuclear Energy, 1983
- A new importance biasing scheme for deep-penetration Monte CarloAnnals of Nuclear Energy, 1982
- A comparative study of different tracklength biassing schemesAnnals of Nuclear Energy, 1980
- Status of Monte Carlo at Los AlamosPublished by Office of Scientific and Technical Information (OSTI) ,1980
- Prediction of Statistical Error and Optimization of Biased Monte Carlo Transport CalculationsNuclear Science and Engineering, 1979
- Optimal Choice of Parameters for Exponential Biasing in Monte CarloNuclear Science and Engineering, 1979
- The Use of Self-Optimized Exponential Biasing in Obtaining Monte Carlo Estimates of Transmission ProbabilitiesNuclear Science and Engineering, 1968