%0 Journal Article %T 修正共轭投影梯度滤子法<br>A Modified Conjugate Projection Gradient Filter Method %A 王祥玲 %A 左双勇 %A 朱志斌< %A br> %A WANG Xiang-ling %A ZUO Shuang-yong %A ZHU Zhi-bin %J 西南大学学报(自然科学版) %D 2018 %R 10.13718/j.cnki.xdzk.2018.11.012 %X 利用共轭投影梯度技术,结合滤子算法的思想,通过修正搜索方向,建立了一个新的共轭投影梯度滤子算法.该算法不需要求解二次规划子问题,而且能有效避免常规滤子算法中的恢复算法.在适当的条件下,证明了算法的全局收敛性.<br>A new conjugate projection gradient filter algorithm is established by modifying the search direction. In this algorithm, conjugate projection gradient technology and filter method are combined. By the introduction of the filter, this algorithm does not need to solve a QP sub-problem. With the idea of the conjugate projection gradient, this method is effective to avoid the restoration algorithm in general filter algorithms. Under some conditions, its global convergence is obtained %K 共轭投影梯度 %K 滤子 %K 非线性规划 %K 全局收敛性< %K br> %K conjugate projection gradient %K filter %K nonlinear programming %K global convergence %U http://xbgjxt.swu.edu.cn/jsuns/html/jsuns/2018/11/20181111.htm