Projection procedures for the nonlocal improvement of linearly controlled processes

From MaRDI portal
Publication:2387066





The author develops a new procedure for the nonlocal improvement of control for the linear state optimal control problem \[ \varphi(x(t_1))+ \int_T \{ F_0(x(t),t)+ \langle F_1(x(t),t), u(t)\rangle \} dt \to \min_{u \in V} \] subject to \[ {\dot x}(t)=b(x(t),t)+A(x(t),t) u(t),\;x(t_0)=x^0,\;u(t) \in U,\;t \in T=[t_0,t_1]. \] The matrix function \(A(x,t)\), the vector functions \(b(x,t)\), \(F_1(x,t)\) and the scalar functions \(\varphi(x)\), \(F_0(x,t)\) are quadratic in \(x\) with continuous coefficients. The set of admissible controls \(u(t), t \in T\), is the class \(V\) of piecewise continuous functions with the values in the closed convex set \(U \subset R^m\).











This page was built for publication: Projection procedures for the nonlocal improvement of linearly controlled processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2387066)