【摘要】 为解决当前气化炉供料系统风险分析不完善的状况,提出1种基于贝叶斯网络和HAZOP的风险分析模型。以某单日投煤量3 000 t级气化炉煤化工企业实际运行情况为研究对象,应用HAZOP法对其进行风险分析,并将HAZOP分析结果中各偏差产生原因转化为贝叶斯网络节点;考虑到先验知识的缺乏问题,引入Leaky Noisy OR模型,通过文献资料和相关领域专家经验知识获得先验概率,并利用贝叶斯网络进行风险分析,找出系统运行的薄弱环节。结果表明:未知因素影响会使各节点的后验概率值差异性降低,更加贴合实际;在引入未知因素影响后,系统运行薄弱环节并未发生改变。更多还原

【Abstract】 In order to solve the imperfect status of risk analysis on the feeding system of gasifier at present,a risk analysis model based on the Bayesian network and HAZOP analysis was proposed.The actual operation situation of the gasifier with a daily coal input of 3000 t level in a coal chemical enterprise was taken as the research object,then the HAZOP method was applied to conduct the risk analysis,and the causes of each deviation in the HAZOP analysis results were converted into the nodes of Bayesian network.Considering the lacking of prior knowledge,the Leaky Noisy OR model was introduced,and the prior probability was obtained through the literature and experts’ experience knowledge in the related fields.The risk analysis was conducted by using the Bayesian network,and the weak links in the system operation were found out.The results showed that the influence of unknown factors could reduce the difference in posterior probability values of each nodes,but it was more realistic.After introducing into the influence of unknown factors,the weak links of system operation did not change.更多还原