A multipoint boundary value problem for systems of linear generalized differential equations with singularities (Q471404)

The paper is concerned with linear integral equations (also known as generalized differential equations in Kurzweil's sense) of the form \[ x(t)=x(s)+\int_s^t {\mathrm d}[A(\tau)]x(\tau)+f(t)-f(s),\eqno(1) \] where \(x\), \(f\) take values in \(\mathbb R^n\) and \(A\) is a matrix-valued function. Given a system of points \(t_1,\dots,t_n\in(a,b)\), the author considers the multipoint boundary value problem described by the \(n\) conditions \[ x_i(t_i+)=x_i(t_i)=x_i(t_i-)=0,\quad i\in\{1,\dots,n\}.\eqno(2) \] Eq.~(1) is assumed to hold whenever \([s,t]\subset[a,b]\) and \([s,t]\cap\{t_1,\dots,t_n\}=\emptyset\). When dealing with Eq.~(1), it is usually assumed that \(A\) has bounded variation on \([a,b]\). The present paper focuses on the situation when each diagonal entry \(a_{ii}\) of \(A\) has bounded variation on each interval \([s,t]\subset[a,b]\) not containing the point \(t_i\), but not necessarily on \([a,b]\). On the other hand, the function \(f\) and the off-diagonal entries of \(A\) still have bounded variation on \([a,b]\). The author presents several results concerning the unique solvability of the problem (1)--(2); the precise conditions are too technical to be reproduced here.