一种新的混合共轭梯度算法

A New Hybrid Conjugate Gradient Algorithm

根据现有的共轭梯度算法,提出了一种新的求解无约束优化问题的混合共轭梯度法.在每一步迭代过程中,新算法总是能生成一个充分下降方向.在Wolfe线搜索下,提出的算法具有全局收敛性.数值实验表明该算法具有良好的计算性能.

Based on some previous conjugate gradient algorithms, a new hybrid conjugate gradient algorithm is put forward for solving unconstrained optimization problems. This algorithm always generates a sufficient descent direction at each iteration. Furthermore, we show that the proposed algorithm possesses global convergence under the Wolfe line search. Finally, some numerical results are reported, which also demonstrate that our algorithm possesses good computational performance.