斜井有杆泵数据挖掘系统中的预处理设计

Preprocessing Design in Data Mining System for Sucker-rod Pumping System in Inclined Well

利用数据挖掘可以优化有杆泵系统的生产管理方案,其数据挖掘目标值宜定为单井生产成本。挖掘数据应是包括油层基础数据、井眼轨迹数据、井下流体数据、井下工具数据、井下作业数据、地面设备数据、日常生产数据等7大类数据在内的所有相关数据。在数据预处理设计中,应进行数据的标准化处理、过滤噪声数据、数据单位的规范化处理、表达术语的统一化处理、成本量化处理等工作。

Data mining system can obtain optimal production management schemes, and the goal may be production cost of a single well. The data can be classified into 7 kinds, such as basic data of the formation, data of the hole trajectory, data of the downhole fluid, data of the downhole tools, data of downhole operation, data of surface equipment, and daily production data. Standardization, filtrating,and unified term of data should be designed in preprocessing in data mining system.