一类不可微凸多目标规划解的最优性充分条件

Sufficient Optimality Conditions for a Class of Nondifferentiable Convex Multiobjective Programming

利用C larke广义梯度,广义次微分将B-凸函数进行推广,给出了正则Lipschitz B-凸等函数的概念,在此基础上我们得到了一类不可微凸多目标规划解的最优性充分条件。

B-vex functions are extended by using the Clarke generalized gradient,generalized sub-differentiable.The definition of the regular Lipschitz B-vex functions are given.Under the above generalized convexity assumption,sufficient optimality conditions for nondifferentiable multiobjective programming are obtained.