摘要
对大庆、大港、孤岛、胜利、华北、沙特6种减压渣油的宽、窄馏分实验数据进行回归分析，得到几组减压渣油物性关联式。在此基础上选定减压渣油的特征性质，借助BP模型神经计算的非线性拟合能力，在行为水平上进行建模与仿真。
The regression analysis on the experimental data of wide and narrow fractions from the six vacuum residua of Daqing, Dagang, Gudao, Huabei and Saudi Arabian crudes were carried out and several physical prperty correlation formulae for vacuum residua were obtained . Based on the above analysis, characteristics of vacuum residue were determined. With the aid of non-linear fitting ability of the BP model neural computation, the model was established and simulation was conducted on the behavior level.
出处
《炼油设计》
1997年第4期34-36,共3页
Petroleum Refinery Engineering
关键词
渣油
数据关联
数学模型
模拟仿真
residue, data correlation, mathematical model, simulation, neural network
