论文标题
部分可观测时空混沌系统的无模型预测
Engineering SYK interactions in disordered graphene flakes under realistic experimental conditions
论文作者
论文摘要
储层计算是预测湍流的有力工具,其简单的架构具有处理大型系统的计算效率。然而,其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间,然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用,该系统模拟了湍流的若干特征。我们表明,使用径向基函数作为非线性投影器,即使只有部分观测并且不知道控制方程,也能稳健地捕捉复杂的系统非线性。最后,我们表明,当测量稀疏、不完整且带有噪声,甚至控制方程变得不准确时,我们的网络仍然可以产生相当准确的预测,从而为实际湍流系统的无模型预测铺平了道路。
We model SYK (Sachdev-Ye-Kitaev) interactions in disordered graphene flakes up to 300 000 atoms ($\sim$100 nm in diameter) subjected to an out-of-plane magnetic field $B$ of 5-20 Tesla within the tight-binding formalism. We investigate two sources of disorder: (i) irregularities at the system boundaries, and (ii) bulk vacancies, -- for a combination of which we find conditions which could be favorable for the formation of the phase with SYK features under realistic experimental conditions above the liquid helium temperature.