On Higher-Order State Constraints.

A higher-order state-constrained optimal control problem is considered. A nondegenerate maximum principle is obtained based on the corresponding inward and outward pointing conditions of the higher order. Simple examples demonstrate how the new optimality conditions exclude degenerate sets of Lagrange multipliers, while the conventional approach fails to do so. The conservation law is derived as a consequence of extremality. The connection to the classical form of optimality conditions is investigated.