摘要
研究了基于贝叶斯网络的摩擦学系统状态辨识知识的获取方法。针对摩擦磨损试验数据一般比较齐全且数据量相对比较大的特点,将贝叶斯网络用于摩擦学系统状态辨识的知识获取,通过计算建立属性之间的依赖关系,获得各种依赖关系下的条件概率表,说明依赖关系的强弱,从而获得对机器摩擦学系统状态辨识有指导意义的客观知识,以利于机器摩擦磨损状态辨识的智能化。将贝叶斯网络用于滑动轴承摩擦磨损试验数据的知识获取,获取该摩擦学系统状态辨识的定性及定量知识。
Knowledge acquisition method of tribologieal system condition identification using Bayesian network was pro- posed. For the much complete testing data of friction and wear, Bayesian network was used in tribological system condition identification for knowledge acquisition. The objective knowledge which has guiding significance for tribological system con- dition identification of machines was acquired through computing the dependence relation among different attributes. This is contributive to the intelligent condition identification of machine tribological system. The developed method based on Bayesian network was used in the knowledge acquisition of sliding bear to acquire the qualitative and quantitative knowl- edge of tribological system condition identification.
出处
《润滑与密封》
CAS
CSCD
北大核心
2009年第12期1-4,共4页
Lubrication Engineering
基金
国家自然科学基金项目(50705070)
教育部博士点新教师项目(20070497029)
关键词
贝叶斯网络
摩擦学系统
状态辨识
知识获取
Bayesian network
tribological system
condition identification
knowledge acquisition
