摘要
讨论了传统可靠性分析方法的优点和缺点,简述了贝叶斯网络的优点及其因果推理与诊断推理,详细讨论了桶消元法的步骤。用一个算例说明贝叶斯网络的推理过程,结果表明,贝叶斯网络的双向推理可以有效地识别电子产品的薄弱环节,为进一步提高电子产品的可靠性及提高维修效率提供依据。
The advantages and disadvantages of traditional reliability analysis methods were discussed.The benefits and the causal inference and diagnosis inference of Bayesian networks were introduced briefly.The process of the bucket elimination algorithm was discussed in detail.An example was given to demonstrate how to complete the bi-directional inference.It is shown that the defects of an electronic product can be identified easily using the bi-directional inference and the results can be use as an basis to further improve the reliability and maintenance efficiency of the electronic products.
出处
《电子产品可靠性与环境试验》
2010年第5期13-17,共5页
Electronic Product Reliability and Environmental Testing
关键词
贝叶斯网络
可靠性
桶消元法
Bayesian networks
reliability
bucket elimination algorithm
