摘要
基于在小型固定流化床反应装置上进行催化裂化反应的试验数据 ,利用人工神经网络的方法将蜡油与渣油催化裂化各种产物的产率与原料油的组成、反应条件进行关联 ,建成蜡油与渣油催化裂化反应数学模型。研究结果表明 ,该模型能很好地拟合试验数据 ,预测不同原料油在不同反应条件下催化裂化的产品产率。
Thepurposeofthework isto establishing mathematicalmodelforpredicting catalytic cracking product slates of heavy oils by using artificialneuralnetworks.Experiments were carried on a fixed fluidized bed reactor.The product yields were correlated with the structuralgroup compositions of heavy oilsand theoperating conditionsby using theartificialneuralnetworksforthemodelestablishment. The results showed thatthe proposed modelcan simulate the experimentalyields and predictthe product yields for various heavy oils atdifferentoperating conditions.
出处
《石油炼制与化工》
EI
CAS
CSCD
北大核心
2000年第3期49-53,共5页
Petroleum Processing and Petrochemicals
关键词
催化裂化
神经网络
重油
产率
数学模型
渣油
catalytic cracking
artificialneuralnetworks
heavy oil
yield
mathematicalmodel
