Fourier Series on the Unit Interval

In summary, a Fourier Series on the Unit Interval is a mathematical representation of a periodic function on the interval [0,1]. It is calculated using the Fourier coefficients, which are determined by integrating the original function over one period and solving for the coefficients using the Fourier series formula. The purpose of using a Fourier Series on the Unit Interval is to approximate a periodic function using simpler sine and cosine functions, which can be useful in various applications. However, it has limitations in accurately representing functions that are not periodic or continuous on the interval [0,1]. For non-periodic functions, other methods such as the Fourier Transform should be used.
  • #1
Euge
Gold Member
MHB
POTW Director
2,052
207
Evaluate the Fourier series

$$\frac{1}{\pi^2}\sum_{k = 1}^\infty \frac{\cos 2\pi kx}{k^2}$$ for ##0 \le x \le 1##.
 
  • Like
Likes benorin, julian, I like Serena and 2 others
Physics news on Phys.org
  • #2
The formula equals

[tex]-4 Re \int^x ds \int^s dt \sum_{k=1}^\infty e^{2\pi kti}[/tex]

[tex]=-4 Re \int^x ds \int^s dt \frac{e^{2\pi ti}}{1-e^{2\pi ti}}[/tex]

[tex]=2Re \int^x ds \int^s dt (1-i \cot \pi t )[/tex]

[tex]=x^2+Bx+C[/tex]

[tex]=x(x-1)+\frac{1}{6}[/tex]

[tex]=x^2-x+\frac{1}{6}[/tex]

because for x=0,1 it is ##\zeta(2)/\pi^2=1/6##. Minimum for x=1/2 is ##-\frac{1}{12}## which coincides with -##\eta(2)/\pi^2## using eta function or alternate zeta function. Ref. https://mathworld.wolfram.com/DirichletEtaFunction.html
 
Last edited:
  • Like
Likes benorin
  • #3
We will evaluate the sum:

\begin{align*}
\frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2}
\end{align*}

for ##-\frac{1}{2} \leq x \leq \frac{1}{2}## first and then make the shift ##x \mapsto x - \frac{1}{2}##.

I will express the sum via a complex contour integration. It employs the fact that the function

\begin{align*}
\dfrac{\cos 2 \pi z x}{\sin \pi z}
\end{align*}

has simple poles at all integer values except when ##\cos 2 \pi z x = 0##, as we now verify. First consider the case where ##\cos 2 \pi z x \not= 0## for ##z=n##,

\begin{align*}
\dfrac{\cos 2 \pi z x}{\sin \pi z} & = \dfrac{\cos 2 \pi n x}{\sin \pi [n + (z-n)]}
\nonumber \\
& = \dfrac{\cos 2 \pi n x}{(-1)^n [\pi (z-n) - \frac{1}{3!} \pi^3 (z-n)^3 + \cdots]}
\nonumber \\
& = (-1)^n \dfrac{\cos 2 \pi n x}{(z-n) \pi [1 - \frac{1}{3!} \pi^2 (z-n)^2 + \cdots]}
\nonumber \\
& = (-1)^n \dfrac{\cos 2 \pi n x}{(z-n) \pi} + \cdots
\end{align*}

Now consider the case where ##\cos 2 \pi z x = 0## for some ##z=n##,

\begin{align*}
\left| \lim_{z \rightarrow n} \dfrac{\cos 2 \pi z x}{\sin \pi z} \right| & = \left| \lim_{z \rightarrow n} \dfrac{\frac{d}{dz} \cos 2 \pi z x}{\frac{d}{dz} \sin \pi z} \right|
\nonumber \\
& = \left| \lim_{z \rightarrow n} \dfrac{- 2 \pi x \sin 2 \pi z x}{\pi \cos \pi z} \right| < \infty
\end{align*}

So when ##\cos 2 \pi z x## vanishes for ##z = n##, there is no pole at ##n##. This allows us to write

\begin{align*}
\frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2} = \frac{1}{2 \pi^2 i} \oint_C \dfrac{\cos 2 \pi z x}{z^2 \sin \pi z} dz
\end{align*}

where the contour ##C## is defined in fig (a).

contoursum.jpg


We have

\begin{align*}
2 \frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2} & = \frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2} + \frac{1}{\pi^2} \sum_{k=-1}^{-\infty} (-1)^k\frac{\cos 2 \pi k x}{k^2}
\nonumber \\
& = \frac{1}{2 \pi^2 i} \oint_{C+C'} \dfrac{\cos 2 \pi z x}{z^2 \sin \pi z} dz
\end{align*}

where the contour ##C'## is defined in fig (a). We complete the path of integration along semicircles at infinity (see fig (b)) since the integrand vanishes there. Since the resulting enclosed area contains no singularities except at ##z=0##, we can shrink this contour down to an infinitesimal circle ##C_0## around the origin (see fig (c)). So that

\begin{align*}
\frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2} = \frac{1}{4 \pi^2 i} \oint_{C_0} \dfrac{\cos 2 \pi z x}{z^2 \sin \pi z} dz
\end{align*}

We expand the integrand in powers of ##z## about ##z=0## and isolate the ##z^{-1}## term. We get

\begin{align*}
\dfrac{\cos 2 \pi z x}{z^2 \sin \pi z} & = \dfrac{1 - \frac{1}{2!} 2^2 \pi^2 x^2 z^2 + \cdots}{z^2 [\pi z - \frac{1}{3!} \pi^3 z^3 + \cdots]}
\nonumber \\
& = \dfrac{1 - 2 \pi^2 x^2 z^2 + \cdots}{z^3 \pi [1 - \frac{1}{6} \pi^2 z^2 + \cdots]}
\nonumber \\
& = \dfrac{1 - 2 \pi^2 x^2 z^2 + \cdots}{z^3 \pi} (1 + \frac{1}{6} \pi^2 z^2 + \cdots)
\nonumber \\
& = \cdots + \left( \dfrac{1}{6} - 2 x^2 \right) \pi \frac{1}{z} + \cdots
\end{align*}

So that for ##- \frac{1}{2} \leq x \leq \frac{1}{2}##,

\begin{align*}
\frac{1}{\pi^2} \sum_{k=1}^\infty (-1)^k \frac{\cos 2 \pi k x}{k^2} & = \frac{1}{4 \pi^2 i} \oint_{C_0} \dfrac{\cos 2 \pi z x}{z^2 \sin \pi z} dz
\nonumber \\
& = \frac{1}{4 \pi^2 i} (-2 \pi i) \left( \dfrac{1}{6} - 2 x^2 \right) \pi
\nonumber \\
& = x^2 - \frac{1}{12}
\end{align*}

Finally, we do the shift ##x \mapsto x - \frac{1}{2}## and obtain:

\begin{align*}
\frac{1}{\pi^2} \sum_{k=1}^\infty \frac{\cos 2 \pi k x}{k^2} = \left( x - \frac{1}{2} \right)^2 - \frac{1}{12} = x^2 - x + \frac{1}{6} .
\end{align*}

for ##0 \leq x \leq 1##.
 
Last edited:
  • Like
Likes benorin, I like Serena and anuttarasammyak
  • #4
The Fourier cosine series for an even periodic function ##f(x)## with period ##2L## is
\begin{array}{cl}f(x)&=\frac{c_0}2 + \sum_{k=1}^\infty c_k \cos\frac{k\pi x}{L}\\
c_k &= \frac 2L\int_0^L f(x) \cos\frac{k\pi x}{L}\,dx\end{array}
Pick ##L=\frac 12## and let ##f(x)## be the desired function. It follows that for ##k>0##: $$\pi^2 c_k=4\pi^2\int_0^{1/2} f(x) \cos(2k\pi x)\,dx=\frac 1{k^2}$$
We have for ##k>0##:
$$\begin{cases}4\pi^2\int_0^{1/2} x \cos(2k\pi x)\,dx = \frac{(-1)^k-1}{k^2}\\
4\pi^2\int_0^{1/2} x^2 \cos(2k\pi x)\,dx = \frac{(-1)^k}{k^2}\end{cases}\implies
4\pi^2\int_0^{1/2} (x^2-x) \cos(2k\pi x)\,dx = \frac{1}{k^2}
$$
Furthermore, ##f(0)=\frac 1{\pi^2}\sum_{k=1}^\infty \frac 1{k^2}=\frac 16##.
So the function is ##f(x)=x^2-x+\frac 16## on the interval [0,1], which is indeed an even periodic function if we extend it periodically.
 
Last edited:
  • Like
Likes benorin

1. What is a Fourier series on the unit interval?

A Fourier series on the unit interval is a mathematical tool used to represent a periodic function as a sum of sine and cosine functions. It is defined on the interval [0,1] and is used in many fields of science and engineering to analyze and approximate periodic phenomena.

2. How is a Fourier series on the unit interval calculated?

A Fourier series on the unit interval is calculated by finding the coefficients of the sine and cosine functions that best fit the given function. These coefficients can be found using integration techniques or by using specialized formulas for specific types of functions.

3. What are the applications of Fourier series on the unit interval?

Fourier series on the unit interval have many applications in science and engineering, such as signal processing, image analysis, and solving differential equations. They are also used in fields such as physics, chemistry, and economics to model and analyze periodic phenomena.

4. What is the difference between a Fourier series on the unit interval and a Fourier series on the real line?

The main difference between a Fourier series on the unit interval and a Fourier series on the real line is the domain on which they are defined. A Fourier series on the unit interval is only defined on the interval [0,1], while a Fourier series on the real line is defined on the entire real number line. This difference affects the types of functions that can be represented and the methods used to calculate the series.

5. How accurate is a Fourier series on the unit interval in approximating a function?

The accuracy of a Fourier series on the unit interval depends on the complexity of the function being approximated and the number of terms used in the series. In general, the more terms included in the series, the closer the approximation will be to the original function. However, there are cases where a Fourier series may not converge or may converge slowly, resulting in a less accurate approximation.

Similar threads

  • Math POTW for University Students
Replies
1
Views
457
  • Calculus and Beyond Homework Help
Replies
3
Views
284
  • Calculus and Beyond Homework Help
Replies
1
Views
215
  • Calculus and Beyond Homework Help
Replies
1
Views
342
  • Math POTW for University Students
Replies
8
Views
648
  • Topology and Analysis
Replies
4
Views
273
  • Differential Equations
Replies
4
Views
2K
  • Calculus and Beyond Homework Help
Replies
6
Views
389
Replies
4
Views
287
  • Calculus and Beyond Homework Help
Replies
3
Views
415
Back
Top