NLP问题解的二阶条件和Lipschitz连续性

The Second-Order Optimality Conditions and Lipschitzian Property of Solution for a Class of Nonlinear Programs

假定所讨论的数学规划问题其函数连续可微且有Lipschitz连续的梯度函数运用Clarke广义Jacobi矩阵,给出了非线性规划(NLP)问题解的二阶最优性必要条件二阶最优性充分条件及非线性参数规划问题解的Lipschitz连续性质,推广了王金德Fiacco的主要结果。

It is assumed that functions in mathematical programs arc continuously differentiable and have Lipschitzian continuous gradient functions . By use of Clarke's generalized Jacobian matrix, the second -order efficient optimality condition, second-order sufficient optimality condition of minimizing solution for nonlinear programs , and the Lipschitzian continuous property of local minimizing solution for nonlinear parametric programs are given . Then the chief results of Wang Jinde and Fiacco are extended .