李丹阳， 蔡金燕， 杜敏杰， 朱  赛
（军械工程学院光学与电子工程系，河北 石家庄 050003）
摘  要：针对故障诊断中的测试序列优化问题，提出一种改进蚁群算法的解决方法。该方法根据二值属性系统的特点，定义状态集向量及测试向量，将故障测试隔离过程转化为向量的位运算过程，将序列优化问题转化为一种最小代价的动态树构造问题，设计灵活的状态转移规则，并根据动态树的分层结构特点，提出一种分层加权和遗传变异相结合的信息素更新策略，解决这种动态树结构的寻优问题。仿真结果表明：该算法以较高的效率收敛到已知最优解，高效实用，为大规模复杂系统的测试优化问题提供了一条新的解决途径，具有一定的工程应用价值。
关键词：测试序列优化；蚁群算法；二值属性系统；动态树
中图分类号：TP277；TP391.9；TM571.62；O212.6      文献标志码：A      文章编号：1674-5124（2013）04-0105-04
Test sequencing optimization based on improved Ant Algorithm
LI Dan-yang， CAI Jin-yan， DU Min-Jie， Zhu Sai
（Department of Electronic and Optical Engineering，Ordnance Engineering College，Shijiazhuang 050003，China）
Abstract: For solving the problem of test sequencing optimization in fault diagnosis， an improved ant algorithm was presented in this paper. According to the feature of the binary attribute system， state-set vector and test-set vector were defined， the fault testing segregation process was transformed to the process of vector operation and the problem of test sequencing optimization was transformed to construct a dynamic tree with minimum cost. Transfer rule of the ant state was designed and a kind of layered weighted pheromone update mechanism was presented， which combine with the variation in GA and solved the optimization problem of tree-construction. Simulation results show that the algorithm convergences to the optimal solution with high efficiency and provides a new way to solve the test sequencing optimization for large-scale complicated system.
Key words: test sequencing optimization； ant algorithm； binary attribute system； dynamic tree