摘要
滴流床是一类气、液、固三相反应器,在石油化工领域应用非常广泛,尤其在加氢裂化过程中常作为主要的反应器使用。本文分析了在重油加氢反应器的实际生产中,引起其压降随生产时间的增加而逐步抬升现象的主要原因,并在前期模型的基础上通过引入铁元素流经总量作为新输入变量,建立了可针对反应器压降随时间变化规律进行有效预测的后期神经网络压降模型,提出了样本筛选的原则。该模型可用于反应器因压降而停工的日期的预测。
The main causes of pressure drop become higher along operating time in hydro-cracking equipment of a refinery plant(VRDS)reactors was analyzed in this paper.An ANN model which uses total flow rate of iron element as a subjoined input parameterbased on the earlier period model was proposed.This model can predict effectively pressure drop of reactors with operating time,andcan also calculate shutdown date of the reactors caused by raising of pressure drop.For building a precise model,it is also discussedhow to select suitable patterns from raw data.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2004年第3期416-420,共5页
Computers and Applied Chemistry