基于修正拟牛顿方程,利用Goldstein—Levitin—Polyak(GLP)投影技术,建立了求解带凸集约束的优化问题的两阶段步长ZhangH.C.非单调变尺度梯度投影方法,证明了算法的全局收敛性.数值实验表明算法是有效的,适合求解大规模问题.
Based on modified quasi-Newton equation, by combining with Goldstein- Levitin- Polyak (GLP) projection technique, a new Zhang H.C. non-monotone two stages stepsize diagonal sparse variable metric gradient projection method for nonlinear optimization problem is presented. The global convergence properties of the new method are proved. The numerical results show that the new method is effective and is fit to solve large-scale problems.