贝叶斯网络(BN)是一种用于描述变量问不确定性因果关系的图形网络模型,用于不确定性系统建模和推理,处理涉及到预测智能推理、诊断、决策风险及可靠性分析的问题.本文首先对贝叶斯网络做了一个简略的介绍,随后综述了贝叶斯网络近30年的发展及功能扩展,对其在工程技术领域的应用包括故障诊断及可靠性分析等方面做了一个回顾,最后对BN现有的不足和未来的研究趋势做了总结和展望.
Bayesian Network (BN) is a graphical network model for the description of the variables between causal uncertainties. It is built for uncertainty modeling and reasoning, processing related to forecasting, intelligent reasoning, diagnosis, decision making, and risk/ reliability analysis. Firstly, a brief presentation on BN was given. Then the development and function expansion of BN within the recent 30 years was outlined, and BN application in the field of engineering technology including fault diagnosis, reliability analysis etc. was reviewed. Finally BN existing deficiencies and future research trends was summarized and discussed.