基于人工神经网络的原油触变特性参数计算

Calculation of Crude Oil Thixotropy Parameters on the Basis of Artificial Neural Network

基于Levenberg-Marquardt算法的前馈BP网络的原油触变特性参数计算方法,计算触变特性参数精度高,不依赖于具体的函数形式,能有效克服传统计算方法的不足。

A computation method of crude oil thixotropy parameters using feed--forward BP network on the basis of Levenberg--Marquardt algorithm is of high accuracy and not dependent on function form and effectively overcomes the disadvantages of traditional methods.