摘要
本文考察了影响高密度水泥浆流变性能的主要因素,并根据实验结果和工程经验建立原材料、配合比和浆体性能对应的数据库。采用人工神经元网络技术,使用三层BP神经元网络对数据进行简化预测,建立配合比和浆体主要性能之间的对应关系,实现配合比和流变性能之间的双向预测。基于以上算法,采用C#语言编制成软件,具有自适应性强、学习能力强、可以处理大量非线性关系数据的能力,也是人工神经元网络技术在油井水泥浆中的新应用。
The main factors of high density cement slurry rheological properties were researched, and the database of raw materials, mix proportion and slurry performance was established according to the experimental results and engineering experience. Artificial neural network technology was used to simplify data prediction, the corresponding relations between the mix proportion and slurry property was established and the two-way prediction between mix proportion and rheological properties was realized. Based on the above algorithm, software was written using C#. This software has strong adaptability and the ability to learn, can handle a large number of data with nonlinear relationship. It~ a new application of artificial neural network technology in oil well cement.
出处
《硅酸盐通报》
CAS
CSCD
北大核心
2013年第9期1905-1909,共5页
Bulletin of the Chinese Ceramic Society
关键词
人工神经元网络
数据库
水泥浆
配合比
流变性能
artificial neural network
database
cement slurry
mix proportion
rheological property
