求解函数优化问题的两种异步并行算法

求解函数优化问题的两种异步并行算法 Two Asynchronous Parallel Algorithms for Function Optimization 康卓 KANG Zhuo 1 2 3 first-author 李艳 LI Yan 1 2 3 刘溥 LIU Pu 1 2 3 康立山 KANG Li\|shan 1 2 3 陈毓屏 CHEN Yu\|ping 1 2 3 武汉大学计算机学院 武汉大学计算机学院 School of Computer School of Computer 软件工程国家重点实验室,武汉大学计算机学院 软件工程国家重点实验室,武汉大学计算机学院 State Key Laboratory of Software Engineering, Wuhan University State Key Laboratory of Software Engineering, Wuhan University 软件工程国家重点实验室 湖北武汉430072,湖北武汉430072,湖北,武汉,430 软件工程国家重点实验室 湖北武汉430072,湖北武汉430072,湖北,武汉,430 对子空间搜索法 （一类多父体重组搜索策略 ）与群体爬山法相结合的一种随机搜索新算法即郭涛算法的特点进行了分析与实例验证 ,并在此基础上提出两种异步并行算法 ,以适应各种类型的并行与分布计算环境 .以Bum p函数的优化问题为例在超级并行计算机上作了并行数值试验 ,得到了迄今最好的结果 Guo Tao proposed a stochastic search algorithm in his PhD thesis for solving function optimization problems. He combined the sub\|space search method（a general multi\|parent recombination strategy） with the population hill\|climbing method. The former keeps a global search for overall situation, and the latter keeps the convergence of the algorithm. In this paper the characteristics of the algorithm are given and some numerical experiments have been done for demostrating the efficiency of the algorithm. The Guo’s algorithm has been parallelized as asynchronous parallel algorithms for suiting different parallel and distributed computing environments. The Bump problem as the numerical example is solved by a super parallel computer and some best results are obtained. 郭涛算法 异步并行 演化算法 函数优化 Guo’s algorithm function optimization asynchronous parallel algorithm evolutionary computation TP301.6 国家自然科学基金资助 （ 70 0 710 42, 6 0 0 730 43,6 0 1330 10 ） 2002-01-01 2021-04-01 1