Indirect solution of optimal control problems with state variable inequality constraints: finite difference approximation.