通信不确定MPI并行程序的测试数据自动生成方法  

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

作　　者：王荣丽[1] 肖宇虹[1] 侯秀萍[1] 

机构地区：[1]长春工业大学计算机科学与工程学院,长春130012

出　　处：《吉林大学学报（理学版）》2016年第6期1378-1382,共5页Journal of Jilin University:Science Edition

基　　金：吉林省科技厅科技成果转化项目(批准号:20130303010GX)

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.Through the analysis of MPI parallel program which was used to transmit information in uncertain way, we gave the mathematical model of 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.

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

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]