1
$\begingroup$

I understand how to find the stability of fixed points by using $f'(x)$ and inserting the values of the fixed points into that. However, I can't figure out how to tell the stability in systems such as

$$\frac{dx}{dt} = a + \sin(x) + \cos(2x)$$

In situations like this, the value of the fixed points varies as '$a$' changes, which I have no trouble finding, its just defining their stability that I struggle with.

|cite|improve this question
$\endgroup$
  • 1
    $\begingroup$ Your question seems too broad. As a starting point, it seems you meant $a$ is time-varying. So you'll enter the realm of nonlinear time-varying systems whose stability analysis requires specific tools such as Lyapunov method. $\endgroup$ – polfosol May 9 '18 at 16:13
  • 1
    $\begingroup$ In the case of the question I took the system from, a is a parameter unrelated to time t, but I rushed and didn't clarify, my bad, the stress of exams is getting to me! $\endgroup$ – user6046937 May 9 '18 at 17:14
1
$\begingroup$

Let us assume a system of this type

$$\dot{x}(t)=f(x;a)$$

with the equilibrium point $x_\text{eq}$. In this $a$ is a parameter. In this kind of problem, the equilibrium point is a function of the parameter $a$ (called bifurcation parameter in this context).

In order to obtain the equilibrium point(s) it is necessary to solve the nonlinear equation

$$0 = f(x_\text{eq};a)$$ $$\text{here: } 0 = a+\sin x_\text{eq}+1-2\sin^2x_\text{eq}$$

Then the stability of the equilibrium point can be obtained by Lyapunov's indirect method (also known as linearization).

Hence, you will have to determine

$$\left.\dfrac{\partial f}{\partial x}\right|_{x=x_\text{eq}}.$$

If this value is positive then, the equilibrium point is unstable. If it is negative hen the equilibrium point is at least locally asymptotically stable.

|cite|improve this answer
$\endgroup$
  • $\begingroup$ The problem is a bit more complicated than that $\endgroup$ – polfosol May 9 '18 at 16:16
  • 1
    $\begingroup$ @polfosol: I think the OP is doing a bifurcation analysis. And $a$ is the bifurcation parameter. But let us wait for the OP to explain what he really means. $\endgroup$ – MrYouMath May 9 '18 at 16:18
  • $\begingroup$ MrYouMath is right, a is the bifurcation parameter. My bad, I should have clarified. A typical question would be: a) Does this equation give a well-defined vector field on a circle. (in the question it uses theta instead of x) b) Find all fixed points as a varies and sketch the flux diagram. c) Draw the bifurcation diagram and classify the occurring bifurcations as a varies. $\endgroup$ – user6046937 May 9 '18 at 16:33
  • 1
    $\begingroup$ By the way, stability of equilibrium in $\mathbb{R}$ or $\mathbb{S}^1$ can be also simply determined just by looking at signs of $f(x; a)$ near the equilibrium. If only qualitative picture is the matter of interest, this could be enough. $\endgroup$ – Evgeny May 10 '18 at 9:12
  • $\begingroup$ @Evgeny Thank you for the addition:). $\endgroup$ – MrYouMath May 10 '18 at 9:40

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.