摘要
针对多UUT(Unit Under Test)并行测试任务调度与资源配置问题,提出了一种遗传蚁群融合算法.应用遗传蚁群融合算法能快速、准确地寻找到具有最大成本效率的多UUT并行测试资源配置和任务序列.建立了多UUT并行测试任务资源描述的数学模型,分析了多UUT测控资源合并的条件,得出最短并行测试时间基础上的最少资源需求,给出了成本效率的定义,设计了一种满足多UUT并行测试任务调度的基因编码方法和路径选择方案.算法初期利用遗传算法的快速收敛性,为蚁群算法提供初始信息素分布,蚁群算法采用双向收敛的信息素反馈方式,避免了对参数的依赖,减少了局部收敛性,加快了收敛速度.实例表明,该算法能很好地解决多UUT任务资源最优调度与配置问题.
A genetic ant colony algorithm （GA-ACA）, which can be used to find the optimized multi- UUT parallel test tasks array and resources configuration quickly and accurately, is proposed. The mathematic model of multi-UUT parallel test tasks and resources was established. The condition of multi-UUT resources combination was analyzed to get minimum resource requirement under minimum test time. The definition of cost efficiency was put which could satisfy multi-UUT parallel test tasks forward. A gene coding and path choosing project scheduling was designed. At the beginning of the algorithm, GA was adopted to provide initialization pheromone for ACO algorithm. To avoid local convergence and parameters dependence, the ACA adopted the dual-convergence pheromone feedback mode. The practice application shows that the algorithm can solve multi-UUT parallel test tasks scheduling and resources configuration problems effectively.
出处
《测试技术学报》
2009年第4期343-349,共7页
Journal of Test and Measurement Technology
基金
总装"十一五"重点预研基金资助项目(51317030103)
关键词
并行测试
遗传蚁群融合算法
成本效率
多UUT
资源配置
任务调度
parallel test
genetic ant colony algorithm （GA-ACA）
cost effieieney
multi-Unit Under Test （UUT）
resource configuration
task scheduling
