Direct computation of the spectral function (original) (raw)
1995, Proceedings of the American Mathematical Society
We would like to find an explicit formula for the spectral function of the following Sturm-Liouville problem: \[ { L f ≡ − d 2 d x 2 f ( x ) + q ( x ) f ( x ) , x ≥ 0 , f ′ ( 0 ) − m f ( 0 ) = 0. \left \{ {\begin {array}{*{20}{c}} {Lf \equiv - \frac {{{d^2}}}{{d{x^2}}}f(x) + q(x)f(x),\quad x \geq 0,} \hfill \\ {f’(0) - mf(0) = 0.} \hfill \\ \end {array} } \right . \] A simple operational calculus argument will help us obtain an explicit formula for the transmutation kernel. The expression of the spectral function is then obtained through the nonlinear integral equation found in the Gelfand-Levitan theory.
Related papers
Mathematical Methods in the Applied Sciences, 2019
For the one‐dimensional Schrödinger equation with short‐range potential on a half‐line x>0, the knowledge of the Jost solution e(ρ,x)∼eiρx, Imρ ≥ 0, x→∞ allows one to solve corresponding spectral problems. In the present work, a new series representation for e(ρ,x) is derived with the aid of the Levin formula for the Jost solution and a recently proposed Fourier‐Laguerre series expansion of the integral kernel from the Levin formula. The representation for e(ρ,x) has the form , where, for the coefficients bn(x), a simple recurrent integration procedure is obtained and the parameter belongs to the unit disk. An analogous representation is derived for the derivative of the Jost solution as well.With the aid of the series representations, numerical solution of the classical spectral problem on the half‐line becomes an easy task. Indeed, computation of the eigenvalues reduces to finding zeros of a polynomial for z∈(−1,1). For computing corresponding normalizing constants, a simple fo...
Mathematical Methods in the Applied Sciences, 2019
The inverse Sturm‐Liouville problem on a half‐line is considered. With the aid of a Fourier‐Legendre series representation of the transmutation integral kernel and the Gel'fand‐Levitan equation, the numerical solution of the problem is reduced to a system of linear algebraic equations. The potential q is recovered from the first coefficient of the Fourier‐Legendre series. The resulting numerical method is direct and simple. The results of the numerical experiments are presented.
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.