Abstract: In order to solve the dynamic failure problem of pulverized coal pressurized conveying system, a reliability evaluation method based on dynamic bayesian network is proposed in this paper.Firstly, based on the analysis of the structure and function of the pulverized coal pressurized conveying system, the fault tree model is constructed, and the reliability analysis model of the pulverized coal pressurized conveying system is established according to the rules of the trans⁃formation from the fault tree model to the dynamic bayesian network.Then, by introducing maintenance factors and time series, using the two⁃way reasoning ability of dynamic bayesian network, the weak links and the dynamic variation law of reliability of the pulverized coal conveying system are determined.The results show that the method can effectively describe the dynamic characteristics of the pulverized coal pressurized conveying system, and the analysis results are more in line with the actual working conditions.

Key words: Pulverized coal pressurized conveying system, Fault tree model, Dynamic bayesian network, Maintenance factor, Reliability analysis

摘要： 为解决煤粉加压输送系统动态失效问题，提出一种基于动态贝叶斯网络的可靠性评估方法。首先，在分析煤粉加压输送系统的结构与功能的基础上构造事故树模型，并根据事故树模型向动态贝叶斯网络转化的规则，建立煤粉加压输送系统的可靠性分析模型。然后，引入维修因素、时间序列，利用动态贝叶斯网络的两向推理能力，确定煤粉加压输送系统的薄弱环节和可靠度的动态变化规律。结果表明，该方法可有效描述煤粉加压输送系统的动态特性，分析结果符合实际工况。

关键词: 煤粉加压输送系统, 事故树模型, 动态贝叶斯网络, 维修因素, 可靠性分析