摘要
首先给出一个基本假设, 假定在多值因果图中, 原因节点不直接影响结果节点, 而是只贡献概率值, 结果节点的状态由这些概率值之和形成的概率分布随机决定. 接着, 通过引入归一化常数来解决推理过程中的归一性问题. 最后, 假定指向同一节点的各连接事件的各状态之间彼此互斥, 这并不会影响推理结果, 这一特性保证了推理过程中任一节点的状态之间彼此互斥, 也简化了推理算法.*
Firstly, this paper proposes an assumption in which the cause nodes do not affect the result nodes directly, and they only contribute an intensity to the probability distribution of the result nodes, and the status of result node is determined randomly by the accumulation of the contributed probabilities of the nodes involved. Secondly, it introduces the concept of unitizing coefficient to provide the unitizing capability during reasoning process. Finally, it points out that there is a property that we can obtain the same inference consequence supposing that they are mutex among all states of all linkage events pointing to the same node in the multi value causality diagram under the assumption above. The property can ensure all states of any node are mutex one another during reasoning process and hence the reasoning algorithm is simplified.
出处
《信息与控制》
CSCD
北大核心
2005年第2期133-136,共4页
Information and Control
基金
博士点基金资助项目(99061116)
重庆市科委攻关计划资助项目(5990)
关键词
多值因果图推理算法
动态因果图
不确定性推理
故障诊断
reasoning algorithm in multi value causality diagram
dynamic causality diagram
probabilistic reasoning under uncertainty
fault diagnosis
