摘要
在汽车发动机故障诊断领域,由于设备内部的复杂性和导致故障的不确定因素,使得解决不确定性问题成为目前发动机故障诊断的首要问题;文章提出了一种用于解决不确定性问题的贝叶斯网络模型,该模型的网络结构学习采用了基于簇的搜索算法;为了获得更高准确率的故障诊断结果,模型加入了对当前信息集的采用,进行结构和参数的在线学习,改进了网络结构,网络通过概率传播算法,推理出产生故障的原因节点;在实例中表明,该模型能准确有效地解决发动机故障诊断中存在的不确定性问题,并与专家系统故障诊断模型做出比较,验证了基于该算法的贝叶斯网络模型在信息不确定性条件下能够提高诊断的准确率。
In motor fault diagnosis fields, because of the complexity of the equipment and the uncertain factor of fault causation, the most important problem of solving motor fault diagnosis is to solve the uncertain problem. The paper contrives a Bayesian network model for solving the aneertain problem, and this network model adopts a search arithmetic based on clusters. In order to get more exact diagnosis result, this model proposes an on-line structure and parameter learning by using current information, which improves the network structure. Through the probability spread algorithm, the Bayesian network can obtain the causation nodes that lead fault. The example makes know that this model can solve the uncertain problem on motor fault diagnosis availably and exactly. Compared with the expert system fault diagnosis, confirms this Bayesian network model based on the paper＇s algorithm can improve the veracity on uncertain information.
出处
《计算机测量与控制》
CSCD
北大核心
2009年第5期830-832,856,共4页
Computer Measurement &Control
关键词
发动机故障诊断
贝叶斯网络
不确定性推理
簇
motor fault diagnosis
Bayesian network
uncertainty reasoning
cluster
