AN EXTENDED MAXIMUM PRINCIPLE (Q4344865)

In optimal control problems, where the operator describing the dynamics of the process and the terminal condition is not regular, that is, its first Fréchet derivative is not surjective at the optimal process, the fundamental first-order necessary conditions offered by the Pontryagin maximum principle do not provide useful and sufficient information. The main goal of the present paper is to obtain first- and second-order necessary conditions, in such abnormal cases as well, for optimal control problems where, beside dynamics, also state endpoint equality constraints can be present. The weak local variant of the main result is derived using Avakov's generalization of Lyusternik's theorem and the second-order Dubovitskij-Milyutin theory developed in previous papers by the authors for the abnormal setting. The strong version of the maximum principle is obtained from the weak principle with the help of the so-called time transformation trick of Dubovitskij and Milyutin. The main results of the paper unify and generalize several results known in the necessity theory of optimal control problems.