MHB Number of Real Solutions to Cosine System Between 0 and 2π

  • Thread starter Thread starter eddybob123
  • Start date Start date
  • Tags Tags
    Cosine System
Click For Summary
The discussion focuses on finding the number of unordered sets of distinct real numbers {t_1, t_2, t_3, t_4} that satisfy a system of cosine equations. The equations are transformed using the cosine double angle identity, leading to a mapping of cosines represented as x_n = cos(t_n). The periodic nature of the mapping suggests that solutions can be derived from periodic points, specifically through the composition of the function f. The derived solution involves the values x = ±√(1/6), leading to the set {θ, π-θ, π+θ, 2π-θ} as a valid solution. The conclusion is that this appears to be the only unique solution to the problem.
eddybob123
Messages
177
Reaction score
0
How many (unordered) sets of pairwise distinct real numbers $$\{t_1,t_2,t_3,t_4\}$$ all between 0 and $$2\pi$$ are there such that in some order they satisfy the following system:
$$\begin{align*}\cos(2t_1)=4\cos(t_1)\cos(t_2)\\
\cos(2t_2)=4\cos(t_2)\cos(t_3)\\
\cos(2t_3)=4\cos(t_3)\cos(t_4)\\
\cos(2t_4)=4\cos(t_4)\cos(t_1)
\end{align*}$$
 
Mathematics news on Phys.org
eddybob123 said:
How many (unordered) sets of pairwise distinct real numbers $$\{t_1,t_2,t_3,t_4\}$$ all between 0 and $$2\pi$$ are there such that in some order they satisfy the following system:
$$\begin{align*}\cos(2t_1)=4\cos(t_1)\cos(t_2)\\
\cos(2t_2)=4\cos(t_2)\cos(t_3)\\
\cos(2t_3)=4\cos(t_3)\cos(t_4)\\
\cos(2t_4)=4\cos(t_4)\cos(t_1)
\end{align*}$$
Nobody else has tried this one, so I'll throw in my thoughts. First, I used the relation $\cos2\theta = 2\cos^2\theta-1$ to write the equations as $\cos t_{n+1} = \dfrac{2\cos^2t_n - 1}{4\cos t_n}\ (n=1,2,3,4)$ (where $t_5$ is interpreted to mean $t_1$). Next, let $x_n = \cos t_n$. Then $x_{n+1} = f(x_n)$, where $f(x) = \dfrac{2x^2-1}{4x}$. So $x_1$ is a periodic point of this mapping, with period $4$. There is also the requirement that $|x_n| \leqslant 1$ for each $n$.

So we want to find $x$ such that $f(f(f(f(x)))) = x$. The composition of four repetitions of the map $f$ seems impossibly complicated, and the best I can do is to form the composition of two repetitions of $f$. In fact, $$f(f(x)) = \frac{2\left(\frac{2x^2-1}{4x}\right)^2-1}{\frac{2x^2-1}{x}},$$ which simplifies to $\dfrac{4x^4 - 12x^2 + 1}{8x(2x^2-1)}.$ if you put that equal to $x$, you get the solutions $x = \pm\sqrt{1/6}$.

Now let $\theta = \arccos\sqrt{1/6}$. Then you can get a solution to the original problem by taking $\{t_1,t_2,t_3,t_4\} = \{\theta, \pi-\theta, \pi+\theta, 2\pi-\theta\}$ (as an unordered set). The trick there is to expand the orbit of period two in the $x$-map to an orbit of period four in the $t$-space by using the fact that there are two numbers in the interval $[0,2\pi]$ taking any given value of the cosine function. In that way, we get four distinct values for the $t_n$ although there are only two distinct values of the $x$'s.

That is the only solution that I can find to this problem, and my guess is that it is unique.
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 4 ·
Replies
4
Views
1K
  • · Replies 1 ·
Replies
1
Views
5K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 26 ·
Replies
26
Views
626
Replies
4
Views
1K
  • · Replies 36 ·
2
Replies
36
Views
7K