Mesh-independence of the Lagrange-Newton method for nonlinear optimal control problems and their discretizations (Q5948172)

The paper is devoted to mesh-independence properties of the Lagrange-Newton method for constrained optimal control problems and their discretizations. The mesh-independence principle, as it was originally formulated by \textit{E. L. Allgower, K. Böhmer, F. A. Potra}, and \textit{W.C. Rheinboldt} [SIAM J. Numer. Anal. 23, 160-169 (1986; Zbl 0591.65043)] for operator equations, had been generalized by the author in a previous paper to the case of inclusions. In the present paper, these ideas are combined with an abstract approximation approach by \textit{K. Malanowski, C. Büskens,} and \textit{H. Maurer} [In: Mathematical Programming with Data Perturbations V, ed. A. V. Fiacco, Marcel Dekker, Lect. Notes Pure Appl. Math. 195, 253-284 (1997; Zbl 0883.49025)]. This approximation approach allows to reformulate the stability and consistency conditions for mesh-independency in optimal control discretizations and leads to new conditions which are easier to check. The latter fact is shown by an analysis of Euler's method and of a certain first-order Ritz type approximation for problems with mixed control-state constraints.