Abstract: By taking the propylene polymerization process as the research object，a method of constructing the knowledge base of the abnormal situation self-recovery system was presented. By applying the results of hazard and operability analysis，mechanism model and related accident cases to the propylene polymerization process，a technical basis for condition monitoring，early warning and self-recovery was established. Based on the results of hazard and operability analysis，alarm rules were formulated，the temperature，pressure，flow rate and liquid level of the process could be real-time monitored，and non-monitored variables could be calculated to give in-time alarm，reasons for abnormal situation，consequences，and preventive measures，which may be important for maintaining normal production.

摘要： 以丙烯聚合过程为研究对象，给出了构建丙烯聚合过程非正常工况自愈调控系统知识库的方法。将危险与可操作性分析结果、机理模型以及相关事故案例应用于丙烯聚合过程，为状态监测、预警和自愈调控等技术应用提供依据。基于危险与可操作性分析结果编写报警规则，实时在线监测过程中温度、压力、流量、组分和液位等变化，计算非可监测变量，在非正常工况时及时报警，并自动给出非正常原因、不利后果、防范建议和措施，这对于保障石化企业安全生产具有重要意义。