梯度相关条件下求解非线性规划问题算法的收敛性

摘要

We prove that the convergence theorems of general algorithms for solving nonlinear programming problems hold under the condition of gradient correlation, meanwhile θ k= 〈▽f(xk) ,dk 〉 does not satisfy any condition.%证明在梯度相关的条件下,一般的求解非线性规划问题算法的收敛性定理在θ<-▽f(xk),dk>无需满足其它任何条件的前提下仍然成立.