1
$\begingroup$

I'm trying to proof the following, and I'm looking for a verification of my proof. If it is incorrect, I'm looking for some help towards a right proof.

If the omega limit set is $\omega(x_0)=\{x^*\}$, then $\lim_{t\to\infty}\phi(t;x_0)=x^*$

Suppose $\lim_{t\to\infty}\phi(t;x_0)$ exists and equals $x'$. Then $\omega(x_0)=\{x'\}$, so it suffices to proof that the limit exists. Now assume for the sake of contradiction that $\lim_{t\to\infty}\phi(t;x_0)$ does not exist. Because $x^*\in\omega(x_0)$, there exists a sequence $(t_n)_{n\in\mathbb{N}}$ such that $\lim_{n\to\infty}\phi(t_n;x_0)=x^*$. Because $\lim_{t\to\infty}\phi(t;x_0)$ does not exist, there exists a $\delta>0$ such that there is no $T$ such that $\phi(t;x_0)\in B_{\delta}(x^*)$ for all $t\geq T$ (otherwise the limit would exist). Fix this $\delta$. From continuity of the flow and existence of a sequence $(t_n)_{n\in\mathbb{N}}$ satisfying $\lim_{n\to\infty}t_n=\infty$ and $\lim_{n\to\infty}\phi(t_n,x_0)=x^*$, it follows that for an arbitrary $\epsilon>\delta$, there exists a sequence $(s_n)_{n\in\mathbb{N}}$ satisfying $\lim_{n\to\infty}s_n=\infty$ such that for all $n\in\mathbb{N}$, $\phi(s_n;x_0)\in\overline{B_{\epsilon}(x^*)}\setminus B_{\delta}(x^*)$, which is in particular a closed set. Now it follows from Bolzano-Weierstrass that $\left(\phi(s_n;x_0)\right)_{n\in\mathbb{N}}$ has a convergent subsequence $\left(\phi(s_{n_k};x_0)\right)_{k\in\mathbb{N}}$, belonging to $\overline{B_{\epsilon}(x^*)}\setminus B_{\delta}(x^*)$, because this set is closed, and thus the limit value is unequal to $x^*$; thus $\omega(x_0)\neq\{x^*\}$, which gives the desired contradiction.

$\endgroup$
2
$\begingroup$

Your proof is correct.​​​​​​​​​​​​​

Edit, in response to the comment: "Would you have any shorter proof in mind or do you think this is one of the simplest possible?".

This is the simplest proof that I know of. The presentation can be made very easy to read too.


Lemma. If $K$ is sequentially compact and $\phi(t_n;x_0)\in K$ for times $t_n\to\infty$, then $\omega(x_0)\cap K\neq\emptyset$.

Proof. Immediate just from the definitions.

If, by contradiction, $x^*$ is not the limit of $\phi(t;x_0)$, there is $\delta>0$ and $t_n\to\infty$ such that $\phi(t_n;x_0)\in B_\delta(x^*)^c$.

Since $x^*\in\omega(x_0)$, there are times $t'_n\to\infty$ such that $\phi(t'_n;x_0)\in B_{\delta/2}(x^*)$.

By continuity of $\phi(\,\cdot\,;x_0)$, there are times $t''_n\to\infty$ such that $\phi(t''_n;x_0)\in \overline{B_\delta(x^*)}\setminus B_{\delta/2}(x^*)=K$. Then the Lemma implies that $\omega(x_0)$ cannot be just $\{x^*\}$.

$\endgroup$
  • $\begingroup$ Thanks for reading. Would you have any shorter proof in mind or do you think this is one of the simplest possible? $\endgroup$ – Václav Mordvinov Oct 29 '18 at 18:06
  • $\begingroup$ Thank you for your edit. Indeed the idea is the same, but it is way shorther and better readable $\endgroup$ – Václav Mordvinov Oct 31 '18 at 7:06
  • $\begingroup$ @LutzL you are totally right! thanks for spotting the typo $\endgroup$ – Federico Nov 5 '18 at 13:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.