基于H-ABC算法的并行测试任务优化研究

Parallel Task Optimization based on H-ABC Algorithm

并行测试任务调度是缩短多个大型测试进程最终完成时间的有效方法,许多研究人员致力于在不违反技术测试顺序的前提下探索测试任务进度优化。对于本项目的优化问题而言,每个测试任务在可用资源上均有各自的资源依赖关系。因此,保证各测试任务依赖资源无冲突的前提下设计有效的算法,满足既定任务间的既定时序关系,而中国对此类问题鲜有讨论。通过对每个任务编码以及输入既定各测试任务间的时序约束与任务与资源间的依赖关系,设计有效的人工蜂群启发式算法(H-ABC)进行优化,以获得各测试任务在其相应资源上的先后安排,同时使得最终测试项目的完成时间最小化。

Parallel task scheduling is an effective method to shorten the total completion time of mutiple large test processes.Researchers have carried out many works focusing on promoting testing efficency under certain technical sequence.Optimization problem which contains tasks with cross relevance is majored in.However,few studies has been down designing a robust algorithm which can ensure resource supplying of different testing process without conflict,while satisfying certain sequential relationship.Through coding each test step,modeling logical order and determining resource strategy,an effective Hybird artificial bee colony(H-ABC)is designed and used to solve the optimization problem,which shows higher speed and fewer constrains compared to traditional methods.