油品质量漂移时的理论分析与操作指导  

To analyze reasons of oil productqualitative change for theory and guide quality up to standard

作　　者：何小荣[1,2] 李博[1,2] 陈丙珍[1,2] 杜民[1,2] 

机构地区：[1]清华大学化学工程系 [2]中国石化总公司三龙经济开发公司

出　　处：《清华大学学报（自然科学版）》1997年第12期60-63,共4页Journal of Tsinghua University(Science and Technology)

摘　　要：利用人工神经网络技术建立炼油厂催化裂化分馏系统油品质量预测模型，实现了油品质量的实时动态监测。在此基础上，对外界条件的变化引起油品质量的漂移，从理论上进行了分析和计算。提出了引起油品质量漂移的原因和可调节参数的调节量。通过对柴油凝点漂移时的实例应用表明，这种方法能够快速、准确地给出柴油凝点漂移时的原因和中段回流量、中段回流温度的调节值，既可以用于离线指导操作，也可以作为在线控制的设定值，实现在线控制，稳定油品质量、避免质量事故的发生，以提高经济效益。On the basic realizing dynamic estimation of oil product qualities in refinery using quality model created by artificial neural networks, the reasons when the quality is not up standard have been analyzed for theory and regulation quantities of every adjustable parameter have been calculated. It is indicated after using the approach for analyzing reasons of diesel condensation point change that the reasons and regulation quantities of middle flow and temperature can be quickly and accurately proved. They are not only as a setpoint for online control but also offline guiding operation so as to decrease quality accident and increase economic profit in refinery.

关 键 词：动态监测 质量模型 石油产品 催化裂化 分馏 

分 类 号：TE626[石油与天然气工程—油气加工工程]