Optimization of non-analog Monte Carlo games using differential operator sampling
A new approach, based on the differential operator sampling technique, is outlined to estimate the derivatives of variance and efficiency with respect to the biasing parameters, using the same simulation that was constructed to solve the primary problem. An algorithm requiring the first and higher order derivatives of the natural algorithm of the second moment to predict minimum-variance biasing parameters is presented. Equations pertaining to the algorithm are derived and solved numerically for an exponentially transformed one-group slab transmission problem for various slab thicknesses and scattering probabilities. Results indicate that optimisation and non-analog simulations can be achieved so that the present method will be useful in self-learning Monte Carlo schemes.
Bibliographic Reference: Article: Nuclear Science and Engineering
Record Number: 199510979 / Last updated on: 1995-08-10
Original language: en
Available languages: en