青岛大学学报（自然科学版）2023，Vol.36Issue(4)：41-45,53,6.DOI:10.3969/j.issn.1006-1037.2023.04.07
基于切比雪夫距离的支撑点选择算法的并行优化研究
Research of Parallel Optimization of Pivot Selection Algorithm Based on Chebyshev Distance
摘要
Abstract
In the pivot selection algorithm for solving Chebyshev distance,how to quickly determine the strength and weakness of pivot has always been a difficult problem to solve due to the large amount of cal-culation.Therefore,a set of fast pivot optimization strategy with Chebyshev distance as the objective function was proposed.Through parallelized analysis,relatively independent computing tasks were found,and OpenMP was used to parallelize the selection of pivot.In order to reduce the time complexity at the al-gorithm level,the Chebyshev distance was converted into the Manhattan distance,which reduces the over-all calculation amount.The multi-threaded method was used to reconstruct the ordering link of the objec-tive function value as a whole,which avoids the meaningless memory fetching overhead.The experimental results show that the pivot optimization algorithm is a more obvious acceleration effect than the traditional method,and the speedup reaches 174.62,and the data dependence problem of the algorithm is solved.关键词
切比雪夫距离/支撑点选择/并行计算Key words
Chebyshev distance/pivot selection/parallel computing分类
信息技术与安全科学引用本文复制引用
陶顺安,李强,尚小敏,周全,张璁..基于切比雪夫距离的支撑点选择算法的并行优化研究[J].青岛大学学报（自然科学版）,2023,36(4):41-45,53,6.基金项目
山东省自然科学基金面上项目(批准号:ZR201910310143)资助. （批准号:ZR201910310143）
