论文标题
部分可观测时空混沌系统的无模型预测
Viable bounce from non-minimal inflation
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 fundamental difficulty in constructing a viable classical bouncing model is to evade the no-go theorem that states that, simultaneously maintaining the observational bounds on the tensor-to-scalar ratio and the non-Gaussian scalar spectrum is not possible. Furthermore, constructing the bouncing phase leads to numerous instabilities such as gradient, ghost, and so on. Most importantly, the model fails to be an attractor, in general, meaning that the solution heavily depends on the initial conditions, resulting in anisotropic (BKL) instability in the system. In this paper, using conformal transformation, we construct a classical bouncing model from a non-minimal slow-roll inflationary model. As a result of the conformal transformation, we show that the model is free of the above instabilities and that it leads to a smooth transition from bouncing to the traditional reheating scenario. We also look at the dynamical analysis of the system in the presence of a barotropic fluid and discover that there exists a wide range of model parameters that allow the model to avoid the BKL instability, making it a viable alternative to inflationary dynamics.
