A refinement of the maximum principle for state constrained optimal control problems under a regularity conitionдоклад на конференции