基于人工神经网络的原油触变模型研究

Research on the Thixotropy Properties of Super Heavy Oil Based on Neural Networks

主要研究超稠油的初始裂降过程,并用人工神经网络模型对原油剪切应力与剪切时间、剪切速率之间的关系进行描述,用Matlab神经网路工具箱进行求解,结果表明应用人工神经网络方法研究原油触变性能得到比较精确的结果,并且计算过程简单,不依赖于具体的函数形式,为研究原油的流变性能提供了一个新的方法和工具。

The initial cracking down process of heavy oil is studied,the relationship of oil thixotropic shear stress and shear time,shear rate artificial neural networks to its three-parameter model are described and solved and with the Matlab neural network toolbox,the results showed that the method of artificial neural network can obtain more accurate results and simpler procedure,and does not depend on the specific functional form,which provide a new method and tool for the study of the rheological properties of crude oil.