在Hausdorff局部凸拓扑线性空间中,借助切锥引进了超有效广义梯度概念。对于群体多目标决策问题,利用供选方案的超有效数,引进了集值映射的联合超有效解。利用切锥的性质建立了联合超有效解在广义梯度意义下的最优性必要条件,利用超有效解集的性质得到了充分条件。推广了现有文献的相关结论。
In Hausdorff locally convex spaces,superly efficient generalized gradient was introduced with the help of contingent cone.For group multiobjective decision making problems,the joint superly efficient solutions of set-valued maps were introduced by means of superly efficient numbers of alternatives.The optimality necessary condition for the joint superly efficient solutions was established in the sense of generalized gradient by using of properties for contingent cone.Optimality sufficient condition was obtained using the properties for the set of superly efficient solutions.Some relevant conclusions were extended in recent references.