二维随机游走与二维格点上的边渗流是否存在关联？

Hey there! It sounds like you're diving deep into percolation theory for your collaborative research—super exciting stuff! Let me break down some key insights about the critical probability $p_c$ in classical percolation that might help fill the gaps you're hitting.

Core Definition Recap #

First, let’s anchor ourselves to the basics: in classical bond or site percolation on a graph, $p_c$ marks the phase transition threshold with these sharp behaviors:

• When $p < p_c$: Almost surely, all connected clusters are finite—no infinite "percolating" cluster exists.
• When $p > p_c$: Almost surely, exactly one infinite percolating cluster exists (this is the "percolation" regime).
• At $p = p_c$: The system is critical. Clusters can grow arbitrarily large, but no infinite cluster forms in most common lattices (like $\mathbb{Z}^d$ for $d \geq 2$). Criticality is governed by universal scaling laws that describe cluster size distributions and other metrics.

Common Research Focus Areas (That Might Align With Your Question) #

Since you mentioned you’ve done extensive research but haven’t found the answer you need, here are some common angles that might match what you’re exploring:

• Exact $p_c$ calculations: Exact values are only known for a handful of lattices—for example, triangular lattice bond percolation has $p_c = 1/2$, thanks to duality arguments. For most other lattices (like square or cubic), we rely on numerical simulations (Monte Carlo methods) or rigorous upper/lower bounds.
• Universality: Critical exponents (the exponents describing how metrics like cluster size behave near $p_c$) are universal across lattices of the same dimension. This means two different 2D lattices will have the same critical exponents, even if their $p_c$ values differ.
• Percolation on non-lattice graphs: If your work uses general graphs (not just Euclidean lattices), $p_c$ can vary wildly. For example, on a regular infinite tree with degree $k$, $p_c = 1/(k-1)$—crossing this threshold means the tree has an infinite connected component.
• Rigorous foundations: Proving the existence of $p_c$ and its properties requires advanced probability tools like ergodic theory and coupling arguments. For $\mathbb{Z}^d$ with $d \geq 2$, we know $0 < p_c < 1$; for $d=1$, $p_c = 1$ (since you need every bond to be open to form an infinite cluster).

A Quick Note to Refine Your Question #

If you can share a bit more about what specifically you’re stuck on—whether it’s computing $p_c$ for your research’s specific graph, understanding how the phase transition applies to your work, or something else entirely—I can give more targeted, actionable advice!

内容的提问来源于stack exchange，提问作者Ryan Honea