论文标题
部分可观测时空混沌系统的无模型预测
Compression of colloidal monolayers at liquid interfaces: in situ vs. ex situ investigation
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。
The assembly of colloidal particles at liquid/liquid or air/liquid interfaces is a versatile procedure to create microstructured monolayers and study their behavior under compression. When combined with soft and deformable particles such as microgels, compression is used to tune not only the interparticle distance but also the underlying microstructure of the monolayer. So far, the great majority of studies on microgel-laden interfaces is conducted ex situ after transfer to solid substrates, for example, via Langmuir-Blodgett deposition. This type of analysis relies on the stringent assumption that the microstructure is conserved during transfer and subsequent drying. In this work, we couple a Langmuir trough to a custom-built small-angle light scattering setup to monitor colloidal monolayers in situ during compression. By comparing the results with ex situ and in situ microscopy measurements, we conclude that Langmuir-Blodgett deposition can alter the structural properties of the colloidal monolayers significantly.