摘要
本文将广义投影梯度方向移植到约束变尺度算法之中,得到了一类新型算法:广义投影梯度型约束变尺度算法,并成功的使用了Armijo规则.该算法将广义投影类可行方向法和约束变尺度算法的优点溶为一体.并且由于使用了拟下降的概念,算法变得更为灵活.
In this paper we combine the generalized project gradient method with the constrainedvariable method and build a type of new algorithm,the constrained variable metric algorithmusing the generalized project gradient. This algorithm has the advantages of both the generalized project graduation algorithm and the constrained variable metric algorithm. Further,thealgorithm is more flexible because of using the quasi-descend technique.
出处
《应用数学》
CSCD
北大核心
1996年第4期470-474,共5页
Mathematica Applicata
关键词
广义投影梯度
约束变尺度法
非线性规划
Generalized project graduation
Armijo rule
Quasi directive derivative