The parametric least squares technique for double eigenvalue problems (Q804260): Difference between revisions
From MaRDI portal
Changed an Item |
Changed an Item |
||
Property / describes a project that uses | |||
Property / describes a project that uses: LINPACK / rank | |||
Normal rank |
Revision as of 12:45, 28 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | The parametric least squares technique for double eigenvalue problems |
scientific article |
Statements
The parametric least squares technique for double eigenvalue problems (English)
0 references
1991
0 references
The following nonstandard eigenvalue problem is considered: \(u''+(\lambda +\mu f+g)u=0\) in (a,c)\(\subset {\mathbb{R}}\), together with \(u(a)=u(b)=u(c)=0,\) \(a<b<c.\) f and g are given (possibly discontinuous) functions. To find eigenvalue pairs \(\lambda\), \(\mu\) and corresponding eigenfunctions u, the least squares method is applied as follows: With suitable trial functions \(\Phi_ j,\quad \Phi_ j(a)=\Phi_ j(c)=0,\) linear combinations \(u^ n=\sum^{n}_{j=1}c_ j\Phi_ j\) are considered, where two constraints are imposed, namely \(\sum^{n}_{j=1}c^ 2_ j=1\) and \(u^ n(b)=0.\) Then, using the technique of Lagrangian multipliers, approximate eigenvalue pairs and eigenfunctions \(u^ n\) are found iteratively. For an example numerical results are presented. The author claims that also more general (even nonlinear) eigenvalue problems can be treated in this way.
0 references
double eigenvalue problems
0 references
nonstandard eigenvalue problem
0 references
eigenfunctions
0 references
least squares method
0 references
Lagrangian multipliers
0 references
eigenvalue pairs
0 references
numerical results
0 references