0% found this document useful (0 votes)
73 views

Ansatz Method

1. The document discusses solving second order, linear differential equations of the form a2(t)y''(t) + a1(t)y'(t) + a0(t)y(t) = f(t). 2. The general solution strategy is to first find the fundamental system for the homogeneous problem, then bring the full problem into standard form and use the Variation of Parameters formula to find a particular solution. 3. The general solution is the sum of the general solution to the homogeneous problem and the particular solution to the full problem.

Uploaded by

DoXuanNam
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
73 views

Ansatz Method

1. The document discusses solving second order, linear differential equations of the form a2(t)y''(t) + a1(t)y'(t) + a0(t)y(t) = f(t). 2. The general solution strategy is to first find the fundamental system for the homogeneous problem, then bring the full problem into standard form and use the Variation of Parameters formula to find a particular solution. 3. The general solution is the sum of the general solution to the homogeneous problem and the particular solution to the full problem.

Uploaded by

DoXuanNam
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Second order, linear differential equations

a2 (t)y 00 (t) + a1 (t)y 0 (t) + a0 (t)y(t) = f (t)

GENERAL SOLUTION STRATEGY:

1 Compute a FS {y1 , y2 } for the corresponding homogeneous problem (f (t) = 0)

a2 (t)y 00 (t) + a1 (t)y 0 (t) + a0 (t)y(t) = 0

2 Bring the full problem to standard form for a2 (t) 6= 0

a1 (t) 0 a0 (t) f (t)


y 00 (t) + y (t) + y(t) =
a (t) a (t) a (t)
| 2{z } | 2{z } | 2{z }
=:p(t) =:q(t) =:g(t)

and compute the Wronskian W (y1 , y2 )(t) using your result from 1 .
Note: You can check the result for the Wronskian using Abel’s theorem.

3 Find a particular solution via the VoP Variation of Parameters formula.


Z  Z 
y2 (t)g(t) y1 (t)g(t)
yp (t) = −y1 (t) dt + y2 (t) dt
W (y1 , y2 )(t) W (y1 , y2 )(t)

4 The general solution is given by

y(t) = C1 y1 (t) + C2 y2 (t) + yp (t)


| {z } |{z}
general solution of hom. problem particular solution of f ull problem

with C1 , C2 ∈ R.

Remarks:
• The constants C1 , C2 ∈ R can be fixed by adding further constraints for the solution, e.g.
initial conditions y(t0 ) = y0 , y 0 (t0 ) = ỹ0 for some t0 , y0 , ỹ0 ∈ R.
• Finding a fundamental system for the homogeneous problem analytically is, in general, diffi-
cult. There are, however, special cases of coefficients a1 , a2 , a3 where it is indeed possible to
compute them (e.g. the constant-coeff. case, Euler type equations, etc.). Furthermore, in the
constant-coefficient case with specific RHS f it is possible to find a particular solution also
by the “Method of Undetermined Coefficients”.
• With the notation L[y](x) = y 00 (x) + p(x)y 0 (x) + q(x)y(x), we can write the standard form
of the full ODE problem as L[y] = g and the homogeneous problem as L[y] = 0. This is
reminiscent of systems of linear equations Ay = b where A is a given matrix, b is a given
vector and y is the unknown vector. Look up the solution strategy for such problems in
Linear Algebra.
Second order, linear differential equations
– constant coefficients, homogeneous –

ODE characteristic equation

ay00 + by0 + cy = 0 becomes with the ”ansatz” y = eλt aλ2 + bλ + c = 0

→ gives solutions of the form



λ1/2 t −b± b2 −4ac
y=e with λ1/2 = 2a

form of general solution depending on the discriminant b2 − 4ac

(1) b2 − 4ac > 0 : λ1 , λ2 are real and distinct

y(t) = C1 eλ1 t + C2 eλ2 t , C1 , C2 ∈ R


b
(2) b2 − 4ac = 0 : λ = − 2a is a (real) double root

y(t) = (C1 + C2 t)eλt , C1 , C2 ∈ R

(3) b2 − 4ac < 0 : λ1/2 = α ± iβ, α, β ∈ R , C1 , C2 ∈ R complex conjugate root pair

y(t) = eαt [C1 cos(βt) + C2 sin(βt)]

Remarks:

• The real part α of λ gives exponential decay or growth, while the imaginary part β gives
oscillations.

• By defining v = y 0 we can rewrite y 00 + by 0 + cy = 0 as a first order system for the vector


Y = (y, v) given by  
0 0 1
Y (t) = Y (t) .
−c −b
| {z }
=:A

The characteristic equation of the matrix is λ2 + bλ + c = 0 and the solution can be given in
terms of the matrix exponential Y (t) = CeAt where C ∈ R2×2 .
Repeated roots

Understanding why the GS of ay00 + by0 + cy = 0 with b2 − 4ac = 0 is of the form


b
y(t) = (C2 t + C1 )e− 2a t .

b
• One solution is obtained via the characteristic equation: y1 (t) = e− 2a t .
b
• A second solution is obtained by making the ansatz y(t) = v(t)e− 2a t :
 b
00  b
0  b

0 = ay 00 + by 0 + cy = a ve− 2a t + b ve− 2a t + c ve− 2a t
b2
   
00 b 0 b
− 2a b b b
=a v − v + 2 e t
+ b v − v e− 2a t + cve− 2a t
0
a 4a 2a
 
b2  b
= av 00 + (−b + b) v 0 + (− + c) v  e− 2a t

| {z }
=0
| 4a
{z }
=0
b
00 − 2a t
= av e

b
• Since a 6= 0 and e− 2a t 6= 0 the equation for v is given by v 00 = 0, so v(t) = C2 t + C1 .
b b b b
• W (e− 2a t , te− 2a t )(t) 6= 0 → {e− 2a t , te− 2a t } is a FS
b
• y(t) = (C2 t + C1 )e− 2a t is the general solution

NOTE: The presented strategy is a special case of a reduction of order: Suppose you know a solution
y1 6= 0 of the second order equation y 00 + p(t)y 0 + q(t)y = 0. Making the ansatz y(t) = v(t)y1 (t) gives
an equation for v reading y1 v 00 + (2y10 + py1 )v 0 = 0, which is a first order equation for w = v 0 .
Const.- coeff., non-homogeneous ODEs with special RHS f
– Method of Undetermined Coefficients (MoUC) –

ay 00 (t) + by 0 (t) + cy(t) = f (t)

ansatz yp (t) for a particular solution of the non-homogeneous problem

RHS f (t) ANSATZ yp (t)


Pn (t) = an tn + . . . + a0 ts (An tn + . . . + A0 )

Pn (t)ert ts (An tn + . . . + A0 )ert

Pn (t)ert sin(mt), Pn (t)ert cos(mt) ts [(An tn + . . . + A0 )ert sin(mt)


+(Bn tn + . . . + B0 )ert cos(mt)]
The exponent s is the smallest integer ∈ {0, 1, 2} such that the ansatz yp (t) does not contain solutions of
the homogeneous problem.

You might also like