The Two-Dimensional Narrow Escape Problem with Deterministic Initial Condition: Mean Escape Time and Escape location probabilities
Abstract
This paper presents a rigorous asymptotic analysis of the two-dimensional narrow escape problem for a Brownian particle initiated at a deterministic position x. While we address the mean escape time, the main novelty lies in our derivation of the asymptotic escape distribution. We derive explicit asymptotic formulas that characterize the dependence of the escape behavior on the relative configuration of the starting point and the window centers. These results remain valid as the starting position approaches the boundary windows and include refined error estimates. The formulas are obtained with two independent methods, each of them having their own interest. A central feature of this work is the rigorous treatment of the underlying mixed boundary value problems in the absence of a contact angle at the escape window boundaries. By leveraging recent results on the maximal regularity of the gradient of the solutions of such mixed boundary value problems, we ensure the analytical validity of our approach despite the inherent loss of regularity of these solutions at the escape window boundaries. Finally, we apply these general results to provide a full sharp asymptotic equivalents for regimes where window sizes scale at specific relative rates.
Related
- Fully Deterministic Shallow Neural Networks with Localized Activations. Part II: Condition Number Estimates
- The Symmetric Location Problem: a Song of Efficiency and Robustness
- A Two-Stage Stochastic Facility Location Problem with Endogenous Demand Learning
- A deterministic polynomial-time algorithm for the Bilu--Linial signing problem on bipartite graphs
- P-curve: An improved solution for the single-facility location problem in two regions with ℓp- and ℓq-norms
BibTeX
@misc{delande2026twodimensional,
title = {{The Two-Dimensional Narrow Escape Problem with Deterministic Initial Condition: Mean Escape Time and Escape location probabilities}},
author = {Loïs Delande and Tony Lelièvre and Boris Nectoux and Thomas Normand},
howpublished = {HAL},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.32829},
eprint = {2610.32829},
archivePrefix = {arXiv},
url = {https://hal.science/hal-05836546v1},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - The Two-Dimensional Narrow Escape Problem with Deterministic Initial Condition: Mean Escape Time and Escape location probabilities
AU - Loïs Delande
AU - Tony Lelièvre
AU - Boris Nectoux
AU - Thomas Normand
PY - 2026
DA - 2026/10/23
PB - HAL
DO - 10.48550/arxiv.2610.32829
UR - https://hal.science/hal-05836546v1
AB - This paper presents a rigorous asymptotic analysis of the two-dimensional narrow escape problem for a Brownian particle initiated at a deterministic position x. While we address the mean escape time, the main novelty lies in our derivation of the asymptotic escape distribution. We derive explicit asymptotic formulas that characterize the dependence of the escape behavior on the relative configuration of the starting point and the window centers. These results remain valid as the starting position approaches the boundary windows and include refined error estimates. The formulas are obtained with two independent methods, each of them having their own interest. A central feature of this work is the rigorous treatment of the underlying mixed boundary value problems in the absence of a contact angle at the escape window boundaries. By leveraging recent results on the maximal regularity of the gradient of the solutions of such mixed boundary value problems, we ensure the analytical validity of our approach despite the inherent loss of regularity of these solutions at the escape window boundaries. Finally, we apply these general results to provide a full sharp asymptotic equivalents for regimes where window sizes scale at specific relative rates.
ER -