BP神经网络关联乙醇-水-异辛醇(含KF)体系液液平衡数据

Correlation of Liquid-Liquid Equilibrium Data for Ethanol-Water-2-Ethylhexanol Containing KF System Using BP Artificial Neural Network

基于神经网络模型,编写通用计算机程序,对乙醇-水-异辛醇三元含盐体系在25℃下的液液平衡数据进行关联,并与文献中所提出的Hand方程进行对比,结果表明:神经网络模型对上述体系平衡数据的关联精度优于Hand方程。神经网络模型为多元液液平衡计算提供行之有效的工具。

Based on the BP Artificial Neural Networks(ANN),the general-purpose computer program is prepared to correlate the liquid-liquid equilibrium data for ethanol-water-2-ethylhexanol containing KF system at 25℃.Compare with the traditional Hand equation in references,the results show that ANN is more accurate than traditional Hand equation and could be an effective method in liquid-liquid equilibrium calculation.