Experiments with Fourier
Series
Part 2: Fourier Approximations
to a Triangular Wave
Our primary interest is to go
in the opposite direction from Part 1, that is, to break down a periodic
function into its "basic" trigonometric components. Specifically, we will
express a periodic function as a linear combination of the simple trigonometric
functions.
If f(x) is a periodic function of period 2 pi, the Fourier series of f(x) is the infinite series
for which the coefficients are calculated from the integral formulas
and
A partial sum of the Fourier series is called a Fourier approximation to f(x).
Our first example is the "triangular" function obtained
by extending the function
to be a periodic function
of period 2 pi.
-
First we consider the coefficients
of the sine terms. What can you conclude about the coefficients
bk?
-
Now look at the coefficients
ak of the cosine terms. Why can we evaluate these by doubling
the integral over the interval [0,pi]?
-
Calculate
a0, and give a geometrical explanation of this
coefficient.
-
Now consider the general
ak for k > 0. What are the first 10
coefficients?
-
Compare the approximations with
the original function: Plot both the function and the approximation for n
= 1...11. Do the approximations appear to converge to the function? For
which values of x is the convergence fastest? For which values of
x is it slowest?
modules at math.duke.edu