基于相异度的测试优选算法  

Algorithm for Test Optimization Selection Based on Dissimilitude

作　　者：陈圣俭[1] 房金龙[1] 张奎山[1] 王栋[1,2] 

机构地区：[1]装甲兵工程学院控制工程系,北京100072 [2]中国航天员科研训练中心,北京100094

出　　处：《计算机测量与控制》2009年第11期2156-2157,2163,共3页Computer Measurement &Control

基　　金：国家自然科学基金资助项目(60871029)。

摘　　要：测试性日益成为现代电子产品的重要性能指标,必须在系统研制初期进行测试性设计,从而在故障诊断过程中,测试费用最少、诊断时间最短;根据贪婪策略提出了一种基于相异度的测试优选算法;在采用基于动态熵的故障隔离策略选择测试中,当动态熵值相同时,结合相异度进行测试优选;实验结果表明,该算法实现简单,能够在诊断时间最短的情况下,测试费用最少。With testability becoming more and more important performance index in moden electronic devices, testability design must be considered at the beginning of system research in older to make the cost of test and the time of diagnosis least. In this paper, an optimized al- gorithm is proposed based on dissimilitude. In this algorithm, each step of test selection based on dynamic--entropy, combined with opti- mized selection of non--excess tests, the number of test and test point get the minimum. By the examination and analysis, the algorithm is simple and could make the cost of test least in the shorter diagnosis time.

关 键 词：故障诊断 非多余测试 相异度 测试性设计 动态熵 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]