深部地质钻探钻进过程流式大数据分析与动态预处理——以辽宁丹东3000 m科学钻探工程为例

Streaming big data analysis and dynamic pre-processing in deep geological drilling process： A case study on the 3000m scientific drilling project in Dandong， Liaoning province

摘要:

深部地质钻探钻进过程数据价值密度低，传统方法难以在钻中流式大数据条件下有效去除尖峰、毛刺等各类钻进过程数据噪声。本文提出一种深部地质钻探钻进过程流式大数据分析与动态预处理方法，并成功应用于辽宁丹东3000 m科学钻探工程。首先，深入分析过程工艺和数据处理需求，建立深部地质钻探钻进过程流式大数据分析与动态预处理框架结构；然后，运用限幅滤波结合过程数据分布特征、司钻/机长人工操作经验去除过程数据中的离群值；接着，引入滑动窗口策略对流式钻进大数据进行动态处理，在每个窗口中运用Savitzky Golay滤波进一步提升数据质量。仿真实验和工程应用结果验证了本文方法具有很好的工程适用性和有效性。

Abstract:

The data quality in deep geological drilling process is poor， and traditional methods are hard to effectively remove all kinds of data noise such as spikes and burrs. A streaming big data analysis and dynamic pre-processing method for deep geological drilling process was proposed and successfully applied to the 3000m scientific drilling project in Dandong， Liaoning province. Firstly， the process mechanism and requirements of data processing are deeply analyzed， and the framework of streaming big data analysis and dynamic pre-processing in deep geological drilling process is established. After that， the outliers in the process data are removed by limiting filtering combined with the distribution characteristics of the process data and the driller’s manual operation experience. Then， the moving window strategy is introduced to dynamically process the big data of convective drilling， and savitzky Golay filter is used in each window to further improve the data quality. Finally， results of simulations and engineering application show that the proposed method has good engineering applicability and effectiveness.