Known Girsanov and Bismut-Elworthy-Li identities are recast as explicit grid-free Monte Carlo algorithms for coupled Fokker-Planck and Hamilton-Jacobi-Bellman systems, with a neural-network demonstration on a Schrödinger bridge problem.
Passive scalar turbulence in high dimensions
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Exploiting a Lagrangian strategy we present a numerical study for both perturbative and nonperturbative regions of the Kraichnan advection model. The major result is the numerical assessment of the first-order $1/d$-expansion by M. Chertkov, G. Falkovich, I. Kolokolov and V. Lebedev ({\it Phys. Rev. E}, {\bf 52}, 4924 (1995)) for the fourth-order scalar structure function in the limit of high dimensions $d$'s. %Two values of the velocity scaling exponent $\xi$ have been considered: %$\xi=0.8$ and $\xi=0.6$. In the first case, the perturbative regime %takes place at $d\sim 30$, while in the second at $d\sim 25$, %in agreement with the fact that the relevant small parameter %of the theory is $\propto 1/(d (2-\xi))$. In addition to the perturbative results, the behavior of the anomaly for the sixth-order structure functions {\it vs} the velocity scaling exponent, $\xi$, is investigated and the resulting behavior discussed.
fields
math.OC 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On the numerical integration of the Fokker-Planck equation driven by a mechanical force and the Bismut-Elworthy-Li formula
Known Girsanov and Bismut-Elworthy-Li identities are recast as explicit grid-free Monte Carlo algorithms for coupled Fokker-Planck and Hamilton-Jacobi-Bellman systems, with a neural-network demonstration on a Schrödinger bridge problem.