### Current browse context:

math.NA

### Change to browse by:

### References & Citations

# Mathematics > Numerical Analysis

# Title: A cell-based population control of Monte Carlo particles for the global variance reduction for transport equations

(Submitted on 20 Jan 2023)

Abstract: We present a population control method with sampling and regulation steps for Monte Carlo particles involved in the numerical simulation of a transport equation. We recall in the first section the difficulties related to the variance reduction methods in the general framework of transport equations; we continue with a brief presentation of the mathematical tools invoked when solving the radiative transport equations and we focus on the importance of the emission and control of existing Monte Carlo particles.

The next part discusses several novel methods based on the cell-based population control method proposed earlier. To this end, we analyze theoretically two types of splitting: one is conservative in energy (at the particle level) and the other is not. Thanks to these results, a new algorithm is introduced that uses the cell-based population control method and a spatial distribution. A numerical comparison of the different types of splitting is proposed in a simplified framework, then the various algorithms presented are compared against two benchmarks : the propagation of a Marshak wave and the propagation of two waves having different intensity and speed scales. To carry out these last tests, we use the multi-physics code FCI2.

## Submission history

From: Gabriel Turinici [view email]**[v1]**Fri, 20 Jan 2023 14:10:19 GMT (1091kb,D)

Link back to: arXiv, form interface, contact.