arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2304.00650math.PRmath.CO
收藏

双重自由边界铁路-yard 二元模型与环形高斯波动

Doubly Free-Boundary Rail-Yard Dimers and Annular Gaussian Fluctuations

University of Connecticut(康涅狄格大学)

Zhongyang Li

AI总结:

研究双重自由边界条件下的铁路-yard 二元测度,揭示自由边界产生无限反射Cauchy因子,推导出高斯波动定理,证明环形反射图像核的协方差结构。

摘要:

我们研究具有左右自由边界条件的铁路-yard 二元测度。双重自由边界几何产生无限多个反射Cauchy因子,出现在分区函数和高度Laplace观测的精确轮廓公式中。这些因子在空边界模型中不存在,但在确定性和二次渐进行为中仍存在。对于允许的分段周期性权重,我们证明了Laplace变换大数定律。在自然矩变量$x=e^{-nβκ}$下,变换收敛给出缩放后的高度函数极限形状。相关的冻结边界满足以下方程系统:$S_χ(w)^β=e^{-nβκ}$,$\frac{d}{dw}\log S_χ(w)=0$,$S_χ(w):=\mathcal G_χ(w)\prod_{r\ge1}\mathcal F_{u,v,r}(w)$。主要的二次结果是一个关于中心化高度Laplace观测的高斯波动定理。协方差是环形反射图像核$\mathsf K_{LL}(z,w)=\partial_z\partial_w\log\frac{Θ_{\mathfrak q}(z/w)}{Θ_{\mathfrak q}(u^2zw)}$,其中$\mathfrak q=(uv)^2$,有指定的积分解释。因此,两个自由边界不仅改变了确定性极限形状:它们将通常的高斯自由场半平面图像结构替换为Laplace测试类上的环形主函数协方差。作为最终的精确可解性结果,我们构建了一个精确生长图采样器,用于有限反射截断的双重自由边界模型,并证明由此得到的$K$-截断分布随着$K\to\infty$在总变差意义下收敛到完整的双重自由边界Gibbs测度,对于每个固定的铁路-yard图,列数有限。

英文摘要:

We study rail-yard dimer models with free boundaries at both ends. The double free-boundary geometry introduces infinitely many reflected Cauchy factors into partition functions and contour formulas for height Laplace observables. For admissible piecewise periodic weights satisfying the stated separation conditions, we prove a Laplace-transform law of large numbers and obtain a limit shape through weak convergence of slope measures. For $β=1$, under the stated regularity and nonexceptional endpoint conditions, we derive a necessary real double-root condition for the frozen boundary. We also prove Gaussian fluctuations of centered height Laplace observables, with covariance determined by an annular reflected-image kernel expressed through theta functions. Finally, we construct an exact growth-diagram sampler for finite reflected truncations and prove that their laws converge in total variation to the full Gibbs measure on every fixed rail-yard graph with finitely many columns as the truncation level tends to infinity.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑