【摘要】 介绍了运用神经网络评估方法对石化生产装置泄漏风险进行量化评估的主要程序:输入变量的选择、风险指标的确定及总风险评估指标的确定等。以某炼油厂柴油加氢精制装置的应用为例,说明该方法可以指导企业根据最终确定的风险等级制定泄漏检测计划,对高风险部位增加检测的频次,既能有效监控密封点的密封效果,又减少不必要的资源投入。更多还原

【Abstract】 This paper introduces main leak quantitative assessment procedures based neural network in petrochemical units:input data,risk index and total risk index etc.with an example of one diesel oil hydrotreating application,it illustrates that this method can guide enterprise to make out the leak inspection plan,increase inspection frequency for high risk sections based on total risk ranks.It can monitor leak effect,and also reduce the resources investment.更多还原