Computing integrals for trigonometric-polynomial quotients

From Companal
Revision as of 19:11, 18 May 2008 by Vipul (talk | contribs) (1 revision)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Description

Here, we describe the integration procedure for integrals of the form:

PVP(x)sinxQ(x)

The sin may be replaced by cos.

Here, we assume that the polynomial Q does not have any zeros on the real axis (except possibly at the origin).

Solution by semicircular contour integration

Further information: semicircular contour integration method

This is typical, for instance, when the integrand is of the form:

f(x):=sinxQ(x)

where the degree of Q is strictly greater than one, and Q has no real zeros. We use the complex-valued function eiz/Q(z). In this case, Jordan's lemma (or the weaker version, the so-called [semicircular contour theorem]]) tells us that the integral along large semicircles tends to zero, so:

PVsinxQ(x)=Im((2πi)res(f;zj))

where the sum is taken over all the zeros zj of f that lie in the upper half-plane.

Note that we can drop the PV (Cauchy principal value symbol), since the integral is absolutely convergent in this case. Also,if Q has only simple poles, so that we have:

Q(z):=j=1n(zzj)

Then, we get the formula:

sinxQ(x)=2πRe(j=1nsin(zj)kj(zjzk))