The general linear problem of the isomonodromic deformation of Fuchsian systems (Q2473736): Difference between revisions

Let \[ \frac{dy}{dz}=\sum_{i=1}^n\frac{B_{i}(a)}{z-a_i}y \] be a parameterized family of Fuchsian systems of the linear differential equations. It is known that the property of a family of Fuchsian systems to be isomonodromic is equivalent to the property of an overdetermined Pfaffian system in many variables \(dy=\omega y\) to be integrable. In the nonresonance case, the form \(\omega\), up to a gauge transformation independent of \(z\) always has the Schlesinger expression \[ \omega = \omega_s = \sum_{i=1}^n \frac{B_{i}(a)}{d(z-a_i)}. \] But in the general case, the form \(\omega\) can look more complicated, as the author shows: Theorem. Each continuous isomonodromic deformation of a Fuchsian system of rank 2 is determined by the Pfaffian form expressed as \[ \begin{multlined} \omega = \omega_s + \omega_n + \omega_res + \omega_add =\\ \sum_{i=1}^n \frac{B_i(a, s)}{d(z-a_i)} + \sum_{l=1}^n c_l(a, s)da_l + \sum_{i=1}^n \sum_{j=1}^n \sum_{k=1}^{r_i}\frac{\gamma_{ijk}(a, s)}{(z-a_i)^k}\,da_j + \sum_{i=1}^n \sum_{j:a_j\in\widetilde{D}} \sum_{k=1}^{r_i} \frac{\widetilde{\gamma}_{ijk}}{(z-a_i)^k}\,ds_j ,\end{multlined} \] where \(\widetilde{D}\) is the set resonance points with scalar monodromy, \(r_i\) is the maximal \(i\)-resonance of the system, and \(s_j\) is the parameter responsible for the choice and dynamics of the direction of maximal normalization at the point with scalar monodromy. The form is assumed to be given on \(Z\times C^r\), where \(r\) is the number of points of \(\widetilde{D}\) and \[ Z=CP^1\times(D(a_0)\setminus\bigcup_{i\neq j}\{z-a_i=0\}) . \] Let us note, that another approach to construct isomonodromic families of the Fuchsian systems, based on differential Galois theory is possible [see the reviewer, Isomonodromic deformations and differential Galois theory, Ukr. Math. J. 59, No. 8, 1131--1134 (2007; Zbl 1138.34042].