摘要
实验研究糠醛精制润滑油过程中温度、醛油比对精制结果的影响.把油品的四个质量指标(折光率、酸值、碱性氮、苯胺点)虚拟成四个组分,对糠醛精制过程建立平衡衡算关系.用人工神经网络算法建立多变量体系的平衡计算.应用该平衡计算结合原油的质量指标预测一定温度和醛油比下精制油的质量指标和收率,得到满意的结果.
The effectoftem perature and the ratio offurfuralto lubricating oilon the resultoffurfuralrefining process is studied. Experim entaldata show thatthere existsan opti- m um region oftem perature forthe refining. By assum ing the fourquality item s (acid num ber、 basic nitrogen、refractive index and Aniline Point) as four com ponents in oil, the m ass balance calculation ofthose com ponents in tw o equilibrium phases is established. An artificialneural netw orks m ethod is developed for the sim ulation ofequilibrium offurfuralrefining. With ex- traction tem perature、theratio offurfuralto lubricating oiland thequality item sdata ofthegiv- en oil, the refining yield and the quality item s ofpurified oilcan be predicted successfully.
出处
《厦门大学学报（自然科学版）》
CAS
CSCD
北大核心
1999年第6期889-895,共7页
Journal of Xiamen University：Natural Science
关键词
糠醛
精制
润滑油
神经网络
模拟
质量指标
预测
FurfuralRefining, Libricating Oil, ArtificialNeuralNetw orks, Quali- ty item s, Sim ulation
