Controllability of two-point nonlinear boundary-value problems by the numerical-analytic method (Q1827018): Difference between revisions

Exact and approximate controllability problems are studied for the finite-dimensional control system \(y' =A(t)y+B(t)u+g(t)+f(t,y,z,u)\), \(z'=F(t,y,z,u)\) with the two types of conditions: (i) \(y(0)=a\), \(z(0)=b\), \(y(T)=c\), \(z(T)=d\), or, (ii) \(y(0)=a\), \(y(T)=c\), \(Q_1 z(0)+Q_2 z(T)=b\). The functions involved are continuous, \(Q_i\), \(i=1,2\), are given matrices and \(T\) is fixed. The method consists in constructing approximate solutions which converge to the desired solution.