论文部分内容阅读
针对自动测试系统中并行测试任务调度复杂、难以优化的问题,提出了一种Petri网技术和模拟退火遗传算法相结合的任务调度优化算法。首先为并行测试系统建立时间Petri网模型,然后将激发的变迁序列集作为并行测试任务调度路径。为了得到最优路径,引入模拟退火遗传(GASA)算法进行搜索。在搜索过程中,将能激发的变迁序列作为染色体,进行选择、交叉和变异。为了防止算法出现收敛过早,陷入局部最优解的现象,还要对个体进行模拟退火操作,最后得到测试完成时间最短的任务调度序列。
Aiming at the problem of complicated and difficult to optimize parallel test task in automatic test system, a task scheduling optimization algorithm based on Petri net and simulated annealing genetic algorithm is proposed. First, set up a time Petri net model for the parallel test system, and then use the excited transition sequence set as the parallel test task scheduling path. In order to get the optimal path, a simulated annealing genetic (GASA) algorithm is introduced to search. In the search process, will be able to stimulate the change sequence as a chromosome, selection, crossover and mutation. In order to prevent the algorithm from converging prematurely and sinking into the local optimal solution, the individual should be simulated and annealed. Finally, the task scheduling sequence with the shortest test completion time is obtained.