“Higher-order” necessary optimality conditions in a linear maximin problem with coupled variables

For a linear maxmin problem with equality constraints, a theorem is proved that provides a basis for constructing an efficient algorithm for its solution.