Solving Second Order Linear Differential Equations With Taylor Series
Solving Second Order Linear Differential Equations With Taylor Series
1 Problem Statement
There is a large class of differential equations–the so-called linear ones–for which
we can find solutions using the Taylor series method discussed in the lecture.
One such differential equation is
d2 x dx
t + + tx = 0 (∗)
dt2 dt
It is a particular case of the more general Bessel differential equation, and
one solution of it is given by the Bessel function J0 (t) that we saw in Chapter
1. Notice that (*) involves not only the first derivative dx dt but also the sec-
d2 x
ond derivative dt2 . For this reason, it is said to be a second order differential
equation.
In this problem we will content ourselves with finding a relationship (specifi-
cally, a recurrence relation) on the coefficients of a Taylor series expansion about
t = 0 of a solution to our equation. Hence consider the Taylor series
∞
X
x(t) = ck tk
k=0
Substituting this into (*) will give you two conditions. The first one is c1 = 0.
What is the other one?
Note: this problem involves some nontrivial manipulation of indices in sum-
mation notation. Do not get discouraged if it feels more difficult than other
problems: it is!
1
2 Solution
First, let us expand the given Taylor series
∞
X
x(t) = ck tk = c0 t0 + c1 t1 + c2 t2 + c3 t3 + . . . = c0 + c1 t + c2 t2 + c3 t3 + . . .
k=0
2
Substituting the summation constructs into the second-order linear differen-
tial equation (*)
∞
X ∞
X ∞
X
k−2 k−1
t k(k − 1)ck t + kck t +t ck tk = 0
k=2 k=1 k=0
Dividing through by t, which, because of power law, cause the first derivative
summation
tk−1
= tk−1 · t−1 = tk−1−1 = tk−2
t
and rearranging the starting k values so that we can factor them out of the
equation later. Separate equations into like equations because of the hint in the
problem statement
∞
X ∞
X ∞
X
k(k − 1)ck tk−2 + kck tk−2 + ck−2 tk−2 = 0
k=2 k=1 k=2
∞
X ∞
X
kck tk−2 = kck tk−2
k=1 k=2
k 2 ck + ck−2 = 0
Answer
ck−2
ck = −
k2