Direct and inverse spectral problems for a Dirac operator with non-local potential (Q2038167): Difference between revisions

The author studies the selfadjoint Dirac operator on a finite segment with non-local potential \[ \left(D_{\mu} y\right)(x)=\begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix} y^{\prime}(x)+\mu \int_{0}^{\pi}\left(y_{1}(t) v_{1}(t)+y_{2}(t) v_{2}(t)\right) d t v(x), \] where \(\mu \in \mathbb{R}, v(x)=\operatorname{col}\left[v_{1}(x), v_{2}(x)\right] \in L^{2}\left(0, \pi, E^{2}\right)\) and \(v_{1}(x), v_{2}(x)\) are real functions. The domain of \(D_{\mu}\) is defined by \[ \begin{gathered} \mathfrak{D}\left(D_{\mu}\right)=\left\{y(x) \in W_{2}^{1}\left(0, \pi, E^{2}\right): y_{1}(0) \sin \alpha+y_{2}(0) \cos \alpha=0,\right. \\ \left.y_{1}(\pi) \sin \beta+y_{2}(\pi) \cos \beta=0\right\}. \end{gathered} \] Spectra of such operators are described. A method for solving the inverse spectral problem is given, a technique to reconstruct the number \(\mu\) and the function \(v(x)\) from the spectral data. For the case of real \(v(x)\), it is shown that the number \(\mu\) and the non-local potential \(v(x)\) are uniquely defined by two spectra. A complete description of the spectral data of the inverse problem for this class of operators is given.