现代化工流程报警系统优化算法及应用  

Alarm system optimization algorithm and its application in chemical process

作　　者：耿志强[1] 朱群雄[1] 顾祥柏[1] 

机构地区：[1]北京化工大学信息科学与技术学院,北京100029

出　　处：《计算机与应用化学》2005年第3期183-188,共6页Computers and Applied Chemistry

基　　金：国家自然科学基金(29976003);教育部科学技术研究重点项目(01024);中石化科学技术研究开发项目(E03007)

摘　　要：过量的流程报警信息给操作者的决策和诊断带来困难,根据现代化工流程报警的特点,提出基于动态模糊聚类算法对化工流程报警系统进行合理分组,采用模糊差异驱动决策算法对报警等级排序,在线优化流程报警系统,提高报警系统的有效性。结合乙烯裂解炉系统的报警优化验证了提出的优化算法的有效性,对现代化工过程参数报警系统管理和操作优化具有较强的实用性。Overload alarms of modern chemical process give the operators many difficulties to decision and diagnosis. For ensuring the safety production and improving operation, management and optimization of alarm information are challenge work that must be confronted. A new system alarm optimization technique based on dynamic fuzzy clustering (DFC) algorithm is proposed according to the relativity of process measured variables. DFC is adopted to group rationally and cluster for information matrix of alarm variables in order to decrease effectively alarms under safety production. And on the basis of DFC fuzzy difference driving decision (FDDD) algorithm is used to rank the clustering center of process alarm variables based on objective process characteristics. The application of practical ethylene cracking furnace alarm system is verified the efficiency of provided algorithm. The proposed optimization algorithm is useful for alarm system management and operation in chemical process.

关 键 词：现代化工流程 报警系统优化 动态模糊聚类 乙烯裂解炉 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TP391.9[自动化与计算机技术—控制科学与工程]