Wavelet Galerkin Solutions of Ordinary Differential Equations
Wavelet Galerkin Solutions of Ordinary Differential Equations
Wavelet Galerkin Solutions of Ordinary Differential Equations
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
410 V. Mishra and Sabina<br />
1<br />
k<br />
( x)<br />
( 1)<br />
a1k<br />
( 2x<br />
k)<br />
.<br />
<br />
k 2<br />
N<br />
2. <strong>Wavelet</strong> <strong>Galerkin</strong> Technique<br />
<strong>Galerkin</strong> Method. It was Russian engineer V.I. <strong>Galerkin</strong> who proposed a projection<br />
method based on weak form. In it a set <strong>of</strong> test functions are selected such that<br />
residual <strong>of</strong> differential equation becomes orthogonal to test functions [8].<br />
Consider one-dimensional differential equation [12, pp. 451-479]<br />
Lu ( x)<br />
f ( x),<br />
0 x 1<br />
(2.1)<br />
with Dirichlet boundary conditions<br />
u( 0)<br />
a,<br />
u(<br />
1)<br />
b.<br />
f is real valued and continuous functions <strong>of</strong> x on [0,1]. L is a uniformly elliptic<br />
differential operator.<br />
L 2 ([0, 1]) is a Hilbert space with inner product<br />
<br />
Suppose that j <br />
is<br />
1<br />
f , g<br />
f ( x)<br />
g(<br />
x)<br />
dt.<br />
2<br />
C on [0, 1] such that<br />
v j 0)<br />
a,<br />
v j ( 1)<br />
<br />
0<br />
v is a complete orthonormal system for L 2 ([0, 1]) and that every j v<br />
( b.<br />
Select a finite set <strong>of</strong> indices j and consider the subspace<br />
S span v j : j <br />
<br />
Let the approximate solution u s <strong>of</strong> the given equation be<br />
us xk<br />
vk<br />
S.<br />
<br />
k<br />
(2.2)<br />
We would like to determine xk in a way that us behaves as if is a true solution on S,<br />
i.e.<br />
u , v f , v j <br />
(2.3)<br />
L s j<br />
j<br />
such that the boundary conditions u s ( 0)<br />
u s ( 1)<br />
0 are satisfied. Substituting u s in<br />
(2.3),<br />
v , v x<br />
f , v j .<br />
(2.4)<br />
<br />
k<br />
L k j k<br />
j<br />
Let X and Y denote the vectors x k and , ,<br />
k<br />
k f v<br />
k<br />
k<br />
A <br />
a j k j , k<br />
, , where a j,<br />
k Lv<br />
k , v j .<br />
(2.4) reduces to the system <strong>of</strong> linear equations<br />
y and A the matrix