定性推理在潜在电路分析中的应用  

A Better Automated Sneak Identification

作　　者：马清亮[1] 胡昌华[2] 陈新海[1] 

机构地区：[1]西北工业大学航天工程学院,陕西西安710072 [2]第二炮兵工程学院自动控制系,陕西西安710025

出　　处：《西北工业大学学报》2003年第2期168-171,共4页Journal of Northwestern Polytechnical University

基　　金：国家自然科学基金 (6 99310 4 0)

摘　　要：将人工智能中的定性推理技术应用于潜在电路分析 ( Sneak Circuit Analysis SCA)。提出了一种直观的权电阻网络形式的电路定性建模方法 ,给出了相应的仿真算法 ,研究了基于状态的功能标记并将其与定性仿真相结合进行潜在电路分析。最后以某导弹点火控制系统为例进行分析 ,结果证明了所述方法的有效性。To our best knowledge, Price, Snooke, and Landry's paper [4] on sneak identification is still the best in the open literature for offering a certain degree of automation in sneak identification. We believe we can offer a degree of automation better than that of their paper. Our sneak circuit analysis (SCA) is based on qualitative reasoning, which is discussed in detail in section 2. Section 2 uses the known qualitative electrical model in which every edge of a graph has a non negative electrical resistance weight; section 2 then develops a qualitative simulation algorithm to obtain qualitative description of the state of each electrical resistance. Section 3 demonstrates that functional labels can be employed to simplify the results of the qualitative simulation and to provide functional information that is indispensable to SCA. Section 4 gives experimental results of a missile ignition system that utilizes our method of automated sneak identification; these results show preliminarily that our method is indeed better than that of Price, Snooke, and Landry.

关 键 词：潜在电路分析(SCA) 定性推理 定性仿真 功能标记 

分 类 号：TP202.1[自动化与计算机技术—检测技术与自动化装置]