摘要
大型设备自身的复杂性、设备的个体差异及工作环境的影响,使设备的故障知识呈现一定的独特性,获取故障知识变得更为困难,针对上述问题,设计了一种能自动更新故障知识的故障诊断方法。该方法采用贝叶斯技术,融合专家的经验知识与设备自身的独特性故障知识,利用故障决策树对单层次和多层次的设备故障进行诊断;最后,给出一个实例证实此方法在实际工程中的简洁性与有效性。该方法已运用于某大型装备的故障诊断中,取得了良好效果。
It is difficult for experts to obtain fault knowledge for large equipment because of its complexity, speciality and different working environment. In order to overcome this difficulty, this paper designed a kind of diagnosis method based on Bayesian fusion theory, which could update fault knowledge automatically through fusing experts’ knowledge and equipment special fault knowledge. Then how to apply it to the diagnosis process of single and multiple layer failure modes for complex large equipment was showed.At last,validated the conciseness and efficiency of this new diagnosis method in real engineering works by an example.
出处
《计算机应用研究》
CSCD
北大核心
2010年第5期1764-1766,共3页
Application Research of Computers
基金
国家杰出青年科学基金资助项目(70825006)
国家自然科学基金资助项目(70901024)
关键词
贝叶斯融合
故障诊断
决策树
自学习
多层次故障
Bayesian fusion
fault diagnosis
decision tree
auto-study
multilayer fault mode
