原油常压蒸馏过程的人工神经网络建模方法

The Model - Building Method of Crude oil Distillation Process Based on Artificial Neural Network

基于神经网络的函数逼近理论以及多层前馈神经网络对非线性现象的强有力刻画与建模能力,对原油常压蒸馏塔侧线抽出馏份进行了神经网络建模,并用GA-BP算法对所建网络模型进行学习和检验,获得了较为满意的结果,探索了用人工神经网络方法建立油品质量估计模型的现实性与可行性.

Based on the function approximation theory of neural networks and MFNN's powerful model-building capability on nonlinear phenomena,This paper established the model of atmospheric side cut with artificial neurul networks, Learning and detecting the established models with GA-BP algorithm, the results are satisfactory, explore the reality and possibility about setting up the quality estimated model of oil by using the method of artificial neural networks.