文档介绍：基于混合二进制粒子群-遗传算法的
测试优化选择研究*
陈希祥,邱静,刘冠军
(国防科学技术大学机电工程与自动化学院长沙 410073)
摘要:测试优化选择是一个组合优化问题。通过对测试选择的目标和约束条件进行深入分析, 建立了其数学模型, 并提出了一种混合粒子群-遗传算法用于求解满足测试性指标要求的最小完备测试集。该算法将遗传算法中的遗传算子引入到二进制粒子群算法中, 既避免陷入局部最优和早熟收敛现象, 又提高了搜索效率。大量实验证明, 对于测试优化选择问题, 混合粒子群-遗传算法能够快速有效的获得全局最优解。
关键词:测试性设计;测试选择;遗传算法;二进制粒子群算法
中图分类号:TP26 N945 文献标识码:A 国家标准学科分类代码:
Optimal test selection based on hybrid BPSO and GA
Chen Xixiang, Qiu Jing, Liu Guanjun
(College of Mechatronical Engineering and Automation, National University of Defense Technology, Changsha 410073, China)
Abstract:Test selection is one of binatorial optimization problems. Based on deep analysis of objectives and constraints of test selection, a mathematical model is founded, and a hybrid algorithm based on BPSO and GA is proposed to solve the plete test set that satisfies the testability requirements. Through introducing ic operators into the BPSO, the algorithm not only avoids the local optimization and premature convergence, but also improves the searching efficiency. Experiments show that the proposed algorithm is fast and effective to achieve global optimal solution of test optimization selection problems.
Key words:design for testability; test selection; GA; BPSO
1 引言
随着武器装备系统性能的提高和复杂性的日益增加,对其进行故障检测与诊断的难度越来越大,开展测试性设计已是当务之急[1]。系统测试性方案优化设计是测试性设计的重要组成部分。作为测试性方案的重要内容之一,测试优化选择[1-2]是测试性方案优化工作的开始,将关系到整个测试性设计工作的好坏,这一问题越来越受到人们的关注。
测试选择的目的在于: