Dynamical Systems - Chaos: Stability condition for a 2-cycle system

Click For Summary
The discussion focuses on the stability conditions for a 2-cycle system defined by the equation x_{n + 1} = F(x_n) = λ(1 - αx_n^2). The user seeks clarification on the conditions for the existence and stability of 2-cycles, particularly the parameter λ. For a 2-cycle to exist, the fixed points of F(F(x)) must be real, leading to the condition λ > √(3/2). Stability is determined by the derivative condition |F'(x)| < 1 at the fixed points, resulting in the stability range -√(3/2) < λ < √(5/2). The user expresses confusion about the theoretical basis for these conditions.
Master1022
Messages
590
Reaction score
116
Homework Statement
Find the values of the parameter ##\lambda## such that the 2-cycle is stable
Relevant Equations
None
Hi,

(This question is part of the same example as a previous post of mine, but I have a question about a different part of it)

I was looking at a question from an exam for a course I am self-teaching. There is a sub-question which asks us to find the values of a parameter for which the 2-cycle system is stable. However, when looking at the solution I don't understand the condition they have used (and I can't see any reference to it in the materials I am using).

Question & Attempt:
Let us imagine we have a system: ## x_{n + 1} = F(x_n) = \lambda ( 1 - \alpha x_n ^2) ##. Then there are some sub-questions about finding the 2-cycle system and the fixed points of it, before the parts I specifically I don't understand.

(i) Find the 2-cycle system
We can do this by doing ## F(F(x_n)) = \lambda - \alpha \lambda ^3 - 2 \alpha ^2 \lambda ^3 x^2 - \alpha ^3 \lambda ^3 x ^4 ##

(ii) Then we find the fixed points of the 2-cycle system
Then we look for the fixed points by setting ## F(F(x)) = x ## which results in the following:
(\lambda - \alpha \lambda x^2 - x) (1 - \alpha x \lambda - \alpha \lambda ^2 + \alpha ^2 x^2 \lambda ^2) = 0

Then the second term is the condition for ##F(F(x)) = x ##, which I thought was also a fixed point.

This leads to two solutions:
\hat x_{1, 2} = \frac{\alpha \lambda \pm \alpha \lambda \sqrt{ 4 \alpha \lambda ^2 - 3}}{2 \alpha ^2 \lambda ^2}

We are told to use ## \alpha = \frac{1}{2} ## from here onwards.
Thus the solutions reduce to: ## \hat x_{1, 2} = \frac{1}{\lambda} \left( 1 \pm \sqrt{ 2 \lambda ^2 - 3} \right) ##

Below are the parts I don't understand the theory for:
(iii) Find the values of ##\lambda## for which the 2-cycle exists (subject of previous question)

The solution exists when ## 2 \lambda ^2 - 3 > 0 \rightarrow \lambda > \sqrt{\frac{3}{2}} ##

(iv) Find the values of ## \lambda ## for which the 2-cycle is stable

The solution scheme says that we do:
\left| \frac{dF(F(x))}{dx} |_{\hat x_{1, 2}} \right| = \left| F&#039;(\hat x_1) F&#039;(\hat x_2) \right| &lt; 1


and then it arrives at the answer: ## -\sqrt{\frac{3}{2}} < \lambda < \sqrt{\frac{5}{2}} ##

I am not sure where this condition comes from... Any help would be greatly appreciated.
 
Physics news on Phys.org
(iii): The fixed points of F \circ F must be real for a 2-cycle to exist.

(iv): Which do you not understand: why |f&#039;(x)| &lt; 1 is the condition for x_{n+1} = f(x_n) to be stable, or why the chain rule gives (F \circ F)&#039;(x_1) = (F \circ F)&#039;(x_2) = F&#039;(x_1)F&#039;(x_2)?
 
Thanks @pasmith !

pasmith said:
(iii): The fixed points of F \circ F must be real for a 2-cycle to exist.
Sure thing!

pasmith said:
(iv): Which do you not understand: why |f&#039;(x)| &lt; 1 is the condition for x_{n+1} = f(x_n) to be stable, or why the chain rule gives (F \circ F)&#039;(x_1) = (F \circ F)&#039;(x_2) = F&#039;(x_1)F&#039;(x_2)?
Neither unfortunately.
 
First, I tried to show that ##f_n## converges uniformly on ##[0,2\pi]##, which is true since ##f_n \rightarrow 0## for ##n \rightarrow \infty## and ##\sigma_n=\mathrm{sup}\left| \frac{\sin\left(\frac{n^2}{n+\frac 15}x\right)}{n^{x^2-3x+3}} \right| \leq \frac{1}{|n^{x^2-3x+3}|} \leq \frac{1}{n^{\frac 34}}\rightarrow 0##. I can't use neither Leibnitz's test nor Abel's test. For Dirichlet's test I would need to show, that ##\sin\left(\frac{n^2}{n+\frac 15}x \right)## has partialy bounded sums...