arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1301.4374cond-mat.stat-mech
收藏

单个布朗轨迹扩散系数最小二乘估计的分布

Distribution of the least-squares estimators of a single Brownian trajectory diffusion coefficient

Denis Boyer, David S. Dean, Carlos Mejia-Monasterio, Gleb Oshanin

摘要:

在本文中,我们研究的分布函数$P(u_α)$的估计$u_α \sim T^{-1} \int^T_0 \, ω(t) \, {\bf B}^2_{t} \, dt$,优化最小二乘拟合的扩散系数$D_f$的一个单一的$d$维布朗轨迹${\bf B}_{t}$。在这里,我们通过考虑$ω(t) = (t_0 + t)^{-α}$形式的权重函数族进一步追求优化,其中$t_0$是时滞,$α$是任意实数,并寻求估计器最有效地滤除波动的$α$值。我们计算$P(u_α)$精确任意$α$和任意空间维度$d$,并表明,只有$α= 2$的分布$P(u_α)$收敛,作为$ε= t_0/T \to 0$,狄拉克δ函数集中在合奏平均值的估计。这使我们可以得出结论,只有估计与$α= 2$具有遍历属性,使合奏平均扩散系数可以获得任何必要的精度从一个单一的轨迹数据,但在一个渐进的实验分辨率为代价。对于任何$α\neq 2$的分布达到,作为$ε\to 0$,一定的限制形式与有限的方差,这意味着,这样的估计是不是遍历。

英文摘要:

In this paper we study the distribution function $P(u_α)$ of the estimators $u_α \sim T^{-1} \int^T_0 \, ω(t) \, {\bf B}^2_{t} \, dt$, which optimise the least-squares fitting of the diffusion coefficient $D_f$ of a single $d$-dimensional Brownian trajectory ${\bf B}_{t}$. We pursue here the optimisation further by considering a family of weight functions of the form $ω(t) = (t_0 + t)^{-α}$, where $t_0$ is a time lag and $α$ is an arbitrary real number, and seeking such values of $α$ for which the estimators most efficiently filter out the fluctuations. We calculate $P(u_α)$ exactly for arbitrary $α$ and arbitrary spatial dimension $d$, and show that only for $α= 2$ the distribution $P(u_α)$ converges, as $ε= t_0/T \to 0$, to the Dirac delta-function centered at the ensemble average value of the estimator. This allows us to conclude that only the estimators with $α= 2$ possess an ergodic property, so that the ensemble averaged diffusion coefficient can be obtained with any necessary precision from a single trajectory data, but at the expense of a progressively higher experimental resolution. For any $α\neq 2$ the distribution attains, as $ε\to 0$, a certain limiting form with a finite variance, which signifies that such estimators are not ergodic.

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。

↑