数据挖掘技术在录井原油性质判别中的应用

Application of data mining techniques in discriminating the properties of mud logging crude oil

随着数据的爆发式增长和大数据技术的不断发展,数据挖掘技术已成为数据分析和知识发现的重要手段。为拓展数据挖掘技术在录井方向上的系统应用思路,在调研、总结其他领域数据挖掘知识及应用方法的基础上,尝试使用渤海油田地化录井数据构建原油性质判别模型,实现了数据挖掘技术在录井解释中的应用,并在特征工程、模型训练等方面积累了经验。构建原油性质判别模型的过程及模型预测结果表明,使用合理的步骤、参数及算法进行数据挖掘,可有效、快速地进行录井解释,数据挖掘技术值得更深入地研究及开发应用。

With the explosive growth of data and the continuous development of big data technology,data mining technology has become an important means of data analysis and knowledge discovery.In order to expand the systematic application of data mining techniques in mud logging direction,on the basis of investigation and summary of data mining knowledge and application methods in other fields,this paper tries to construct a discriminant model of crude oil properties by using geochemical logging data in Bohai Oilfield,which realizes the application of data mining techniques in mud logging interpretation and has accumulated experience in the aspects of characteristic engineering,model training,etc.The process of constructing the discriminant model of crude oil properties and the prediction results of the model show that mud logging interpretation can be carried out effectively and quickly by using reasonable steps,parameters and algorithms for data mining.Data mining techniques are worth more in-depth research and development applications.