(h,φ)多目标规划的鞍点最优性条件

Saddle Point Optimality Conditions For(h,φ)-Multiobjective Programming

利用Ben-Tal广义代数运算定义了(h,φ)-多目标规划的(h,φ)-K-T鞍点,得到鞍点是有效解的充分条件。当目标函数和约束函数是(h,φ)-η广义凸函数时,在广义约束规格条件下得到鞍点是有效解的必要条件。

A new kind of （h,ψ） - muhiobjective programming saddle points, termed （h,ψ） - K - T saddle points is introduced by using Ben - Tal generalized algebraic operations. We derive that （h,ψ）- K - T saddle points is the sufficiency conditions of efficient solution. When the objective function and constrained function is （h,ψ） -η generalized convex functions, we can derive that （h,ψ） - K - T saddle points are the necessary conditions of efficient solution under the generalized constraint qualification conditions.