摘要
针对模拟电路故障诊断中的测试节点优选问题,首先定义了决策矩阵A及其相关概念,对测试节点优选问题建立了数学模型,将测试节点优选问题转化为典型的0/1规划问题,将测试节点的优选过程变成数学模型的求解过程;然后针对建立的数学模型,提出了一种改进遗传算法,并通过和传统遗传算法、分枝定界法的比较,对算法性能进行了分析。仿真结果表明,提出的节点优选方法可靠有效,具有较高的工程应用价值。
The test node selection problem in analog fault diagnosis was studied. Firstly, the analysis matrix A and the related concepts were proposed, and a new mathematical model for the test node optimization was presented. The issue of test node optimization was transformed to a problem of typical 0/1 integer linear programming, and the optimization of the test node selection was transformed to the solving of the mathematical model. Secondly, aiming at solving the mathematical model, an improved Genetic Algorithm （GA） was proposed and the performance of the algorithm was analyzed by comparing it with the traditional GA and the branch and bound method. Finally, the results of the experiment show that the method proposed in this paper is reliable and effective, which is an adaptable engineering application.
出处
《电光与控制》
北大核心
2013年第4期63-67,共5页
Electronics Optics & Control
基金
河北省重点基础研究项目资助(10963529D)
关键词
故障诊断
决策矩阵
0
1规划
遗传算法
fault diagnosis
analysis matrix
0/1 integer linear programming
GA
