FAULT-DIAGNOSIS WITH CONTINUOUS SYSTEM MODELS

被引：3
|
作者：
CHU, BTB
机构：
[1] Department of Computer Science, University of North Carolina at Charlotte, Charlotte
来源：
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS | 1993年 / 23卷 / 01期
基金：
美国国家科学基金会;
关键词：
D O I：
10.1109/21.214767
中图分类号：
TP3 [计算技术、计算机技术];
学科分类号：
0812 ;
摘要：
Most present research on diagnostic reasoning has dealt with discrete causal relationships. However continuous causal models, particularly regression models, are frequently encountered in many applications, especially in equipment/instrument diagnostic applications. A unified diagnostic reasoning model that deals with both continuous as well as discrete causal relationships is presented. The diagnostic model significantly extends the formal probabilistic diagnostic reasoning model of Peng and Reggia and other works. Statistical theories are used to formally derive conditional causation probabilities based on continuous system models. The derived conditional causation probabilities can be used along with discrete causal relationships provided by experts to find the most probable diagnostic hypothesis for a given set of observations.
引用
收藏
页码：55 / 64
页数：10
相关论文
共 50 条
12345