Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

lOMoARcPSD|9327020

2.5: Deduce Stability of Equilibrium Solutions using Taylor Series Expansions

Kiam Heong Kwa


(Dated: October 5, 2011)

Consider the first order autonomous equation point of f , the Taylor series can be reduced slightly to

f ′ (y 0 ) f ′′ (y 0 )
f (y) = (y − y 0 ) + (y − y 0 )2 + · · ·
dy 1! 2!
= f (y). (1)
dt f (n) (y 0 ) f (n+1) (y 0 )
+ (y − y 0 )n + (y − y 0 )n+1 + · · · . (3)
n! (n + 1)!

Recall that an equilibrium solution of Eq. (1) is any constant Let n be the smallest positive integer such that f (n) (y 0 ) 6=
solution y(t ) of the equation; in particular, y(t ) = y 0 for all 0. (Generally, such an integer n depends on the particular
dy critical point y 0 of f .) This reduces the Taylor series further
t ∈ R, where y 0 is the initial value of y(t ). Since = 0,
dt to
this is equivalent to the condition that f (y 0 ) = 0. This way
one sees that a constant function y(t ) = y 0 is an equilibrium f (n) (y 0 ) f (n+1) (y 0 )
f (y) = (y − y 0 )n + (y − y 0 )n+1 + · · · . (4)
solution of Eq. (1) if and only if f (y 0 ) = 0. n! (n + 1)!

Let y(t ) = y 0 be an equilibrium solution of Eq. (1). The Since for y sufficiently close to y 0 , |y −y 0 |N ≪ |y −y 0 |n for all
(in)stability of y(t ) relies on the local property of f near y 0 . integer N > n, it follows that the local behaviors of f near y 0
To facilitate our discussion, let us suppose that f is smooth f (n) (y 0 )
near y 0 in the sense that it can be differentiated as many are dominated by the leading term (y − y 0 )n on the
n!
times as we want in a neighborhood of y 0 . Specifically, there right-hand side of Eq. (4). In particular, as a consequence of
is a Taylor series expansion Taylor’s theorem, it can be shown that the sign of f (y) equals
the sign of f (n) (y 0 )(y − y 0 )n for all y sufficiently close to y 0 .
As a corollary, one has the following conclusion about the
stability of y(t ) = y 0 as an equilibrium solution of Eq. (1):
f ′ (y 0 ) f ′′ (y 0 )
f (y) = f (y 0 ) + (y − y 0 ) + (y − y 0 )2 + · · ·
1! 2! 1. If n is odd and f (n) (y 0 ) < 0, then y(t ) is asymptotically
(n)
f (y 0 ) f (n+1) (y 0 ) stable.
+ (y − y 0 )n + (y − y 0 )n+1 + · · · (2)
n! (n + 1)!
2. If n is odd and f (n) (y 0 ) > 0, then y(t ) is unstable.

3. If n is even, then y(t ) is semi-stable.


valid in a neighborhood of y 0 . Note that since y 0 is a critical

You might also like