论文标题
部分可观测时空混沌系统的无模型预测
The CIDER simulations: Nonlinear structure formation in the Constrained Interacting Dark Energy Scenario
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 present for the first time a suite of cosmological simulations for a particular class of interacting Dark Energy cosmologies characterised by a background expansion history constrained to be indistinguishable from $Λ$CDM. Such Constrained Interacting Dark Energy scenario -- or CIDER -- has been recently proposed by Barros et al. 2019 and has the appealing feature of suppressing structure formation at late times, thereby possibly alleviating the persisting $σ_{8}$ tension while leaving background observables unaffected. A crucial step to assess the viability of such scenarios is then represented by quantifying their impact on structure formation at non-linear scales, which is what we start investigating with the simulations discussed in the present work. We show that -- for reasonable parameter choices -- the reconstructed scalar potential is close to an exponential for most of the matter dominated epoch, and that the nonlinear evolution of structures in these models imprints specific footprints on matter and halo statistics that may allow to break degeneracies with standard cosmological parameters.
