BP网络与NRTL方程在汽液平衡数据预测中的比较研究

A comparative study between BP neural network and NRTL equation in the prediction of the vapor-liquid phase equilibrium data

用BP神经元网络,对乙醇-环己烷二元体系的汽液相平衡数据进行关联和预测,并用三参数NRTL方程的关联及预测结果进行对比研究.结果表明,BP神经元网络对汽液平衡数据的关联和预测结果在精度上均明显优于NRTL方程。

The vapor-liquid phase equilibrium data of binary system of ethanol- cyclohexane has been correlated and predicted by means of BP neural network. In addition, a comparative study between the correlation of three parameter NRTL equation counteracting and the predicted result has been made. The result shows that BP neural network is obviously superior to NRTL equation in terms of accuracy.