论文标题
部分可观测时空混沌系统的无模型预测
Compact Groups in GDM Cosmological Simulations
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。
In this work, we study some properties of the Hickson Compact Groups (HCGs) using N-body simulations for the Generalized Dark Matter (GDM) model, described by three free functions, the sound speed, the viscosity and the equation of state. We consider three GDM models associated with different values of the free functions to neglect collisional effects. We constructed the initial seeds of our simulations according to the matter power spectrum of GDM linear perturbations, which hold a cut-off at small scales, and explored their effects on the non-linear structure formation at small and intermediate scales. We generated mock catalogues of galaxies for different models and classify HCGs by implementing an algorithm that adapts the original selection method for mock catalogues. Once the HCGs samples are classified, we analyzed their properties and compared them between models. We found that a larger amount of HCGs are counted in GDM simulations in comparison to CDM counts. This difference suggests that HCGs can proliferate within GDM despite the suppressed substructure, which indicates a possible modification in the HCG formation process within models where DM is not perfectly like CDM. Additionally, we identified different mechanisms of clustering, for models with a large amount of galaxy-halos self-agglomerate because of their abundance while models with fewer galaxy-halos need massive halos acting as a dominant potential well. Finally, by comparing distributions of different observables of simulated HCGs against observations, we found a good agreement in the intrinsic properties. However, a discrepancy in the velocity dispersion remains unsolved.