arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2404.00337math.DS
收藏

游荡动力学的多能性

Pluripotency of wandering dynamics

Tokai University(东海大学);Tokyo Metropolitan University(东京都立大学)

Shin Kiriki, Yushi Nakano, Teruhiko Soma

AI总结:

本文提出一种称为“多能性”的扰动机制,使得在均匀双曲集中指定轨道的符号和统计行为,经过任意小扰动后,能在正勒贝格测度集的所有点的前向轨道上实现,从而从统计和几何角度重新编程动力学。

摘要:

本文分离出一种扰动机制,我们称之为“多能性”,通过该机制,均匀双曲集中指定轨道的符号和统计行为,在任意小的扰动后,可以在正勒贝格测度集的所有点的前向轨道上实现。从这个意义上说,多能性提供了一种从统计和几何角度重新编程动力学的方法:正测度集中所有点的经验测度可以渐近地跟随双曲集中指定轨道的经验测度。我们首先给出了一个抽象准则,用符号行程描述表述,它等价于多能性的强形式。然后我们证明,该机制在高维非双曲动力学中稳健地出现。更精确地说,对于每个 $2\le r<\infty$ 和 $\dim M\ge 3$,存在一个 $C^r$-开集的微分同胚,具有野性搅拌器-马蹄,使得该开集中的每个微分同胚对于搅拌器-马蹄的一个稠密不变子集都是强多能的。作为应用,这产生了同一开集内具有非平凡狄拉克物理测度和具有历史游荡域的微分同胚的稠密类,提供了与Takens最后一个问题相关的新机制。

英文摘要:

This paper isolates a perturbative mechanism, which we call \emph{pluripotency}, by which the symbolic and statistical behavior of prescribed orbits in a uniformly hyperbolic set can be realized, after an arbitrarily small perturbation, along the forward orbits of all points in a set of positive Lebesgue measure. In this sense, pluripotency provides a way of reprogramming dynamics from both statistical and geometric viewpoints: the empirical measures of all points in a positive-measure set can be made to asymptotically follow those of a prescribed orbit in the hyperbolic set. We first give an abstract criterion, formulated in terms of symbolic itinerary descriptions, which is equivalent to a strong form of pluripotency. We then prove that this mechanism occurs robustly in higher-dimensional non-hyperbolic dynamics. More precisely, for every $2\le r<\infty$ and $\dim M\ge 3$, there exists a $C^r$-open set of diffeomorphisms with wild blender-horseshoes such that every diffeomorphism in this open set is strongly pluripotent for a dense invariant subset of the blender-horseshoe. As applications, this yields dense classes of diffeomorphisms with non-trivial Dirac physical measures and with historic wandering domains inside the same open set, providing a new mechanism related to Takens' last problem.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑