论文标题
部分可观测时空混沌系统的无模型预测
On the role of mechanosensitive binding dynamics in the pattern formation of active surfaces
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 actin cortex of an animal cell is a thin polymeric layer attached to the inner side of the plasma membrane. It plays a key role in shape regulation and pattern formation on the cellular and tissue scale and, in particular, generates the contractile ring during cell division. Experimental studies showed that the cortex is fluid-like but highly viscous on long time scales with a mechanics that is sensitively regulated by active and passive cross-linker molecules that tune active stress and shear viscosity. Here, we use an established minimal model of active surface dynamics of the cell cortex supplemented with the experimentally motivated feature of mechanosensitivity in cross-linker binding dynamics. Performing linear stability analysis and computer simulations, we show that cross-linker mechanosensitivity significantly enhances the versatility of pattern formation and enables self-organized formation of stable contractile rings.
