吉林大学学报(理学版)

• 计算机科学 • 上一篇    下一篇

通信不确定MPI并行程序的测试数据自动生成方法

王荣丽, 肖宇虹, 侯秀萍   

  1. 长春工业大学 计算机科学与工程学院, 长春 130012
  • 收稿日期:2016-04-12 出版日期:2016-11-26 发布日期:2016-11-29
  • 通讯作者: 侯秀萍 E-mail:houxiuping@mail.ccut.edu.cn

Test Data Automatic Generation Method for Communicationof Uncertainty MPI Parallel Program

WANG Rongli, XIAO Yuhong, HOU Xiuping   

  1. School of Computer Science and Engineering, Changchun University of Technology, Changchun 130012, China
  • Received:2016-04-12 Online:2016-11-26 Published:2016-11-29
  • Contact: HOU Xiuping E-mail:houxiuping@mail.ccut.edu.cn

摘要: 通过对用不确定方式传递信息的MPI并行程序进行分析, 给出测试数据自动生成的数学模型, 提出相似路径概念, 并将交叉思想与协同进化机制融入到人工蜂群算法中, 提出一种测试数据生成算法〖CD2〗交叉协同进化人工蜂群算法. 应用该算法求解并行程序的测试数据, 并与人工蜂群算法、 随机法进行对比分析. 实验结果
表明, 该方法可以求解测试数据, 并降低了时间消耗.

关键词: 交叉协同进化人工蜂群算法, 通信不确定MPI并行程序, 测试数据, 相似路径

Abstract: Through the analysis of MPI parallel program which was used to transmit information in uncertain way, we gave the mathematical modelof automatic generation of test data, put forward the concept of similar path, then integrated crossidea and coevolutionary mechanism into artificial bee colony algorithm, and proposed a test data generation algorithm: cross coevolutionary artificial bee colony algorithm. We used the proposed algorithm to solve
the test data of parallel programs, and compared with the results of artificial bee colony algorithm and random method. Experimental results show that the method can solve the test data and reduce time consumption.

Key words: communication of uncertainty MPI parallel program, similar path; cross coevolutionary artificial bee colony algorithm; test data

中图分类号: 

  • TP311