摘要
介绍了渣油评价中人工神经网络分析方法 ,其目标是以少量的关键参数推演出渣油评价所需的其他参数 ,并试图通过改变网络输入参数来确定不同性质指标对推演渣油其他参数的影响程度。文中给出了人工神经网络方法与多元线性回归方法的对比 ,表明人工神经网络用于模拟减压渣油的评价方法和实验过程的可行性 ,并为该方法用于仿真实验过程指出应用前景。
The ANN method introduced in this article is used to evaluate and analyze heavy oil. It can be used to infer other parameters needed in evaluating heavy oil by using a few pivotal parameters, and can be also be used to conclude the affecting degree that different kind of property affects the quality of heavy oil by changing the input parameters of ANN. This article presents the capacity contrast between ANN and multiple linear regression analysis method, indicates the feasibility that ANN is used in evaluating method of simulating heavy oil and experiment process, and points out the applying prospect of this method being used in emulating experiment.
出处
《计算机与应用化学》
CAS
CSCD
2000年第3期215-218,共4页
Computers and Applied Chemistry
基金
国家自然科学基金! ( 2 93560 0 0 1 )
中国石油化工总公司资助项目
