论文标题
部分可观测时空混沌系统的无模型预测
Semidilute Principle for Gels
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。
Polymer gels such as jellies and soft contact lenses are soft solids consisting of three-dimensional polymer networks swollen with a large amount of solvent. For approximately 80 years, the swelling of polymer gels has been described using the Flory--Huggins mean-field theory. However, this theory is problematic when applied to polymer gels with large solvent contents owing to the significant fluctuations in polymer concentration. In this study, we experimentally demonstrate the superiority of the semidilute scaling law over the mean-field theory for predicting the swelling of polymer gels. Using the semidilute scaling law, we experimentally determine the universal critical exponent $ν$ of the self-avoiding walk via swelling experiments on polymer gels. The experimentally obtained value $ν\simeq 0.589$ is consistent with the previously reported value of $ν\simeq 0.588$, which was obtained by precise numerical calculations. Furthermore, we theoretically derive and experimentally demonstrate a scaling law that governs the equilibrium concentrations. This scaling law contradicts the predictions made by de Gennes' $c^{*}$ theorem. A major deficiency of the $c^*$ theorem is that the network elasticity, which depends on the as-prepared state, is neglected. These findings reveal that the semidilute scaling law is a fundamental principle for accurately predicting and controlling the equilibrium swelling of polymer gels.