基于K均值聚类方法的抽油机井系统能耗分析

Energy Consumption Analysis of Rod Pumping Well System Based on K-Means Clustering

在智慧油田的助力下,大多数油田已基本实现了油田生产数据的自动化采集,然而这些数量巨大、维度高的数据信息并未得到有效利用。本文将K均值聚类分析方法应用到抽油机井能耗数据分析中,进行了能耗数据异常识别和油井能耗分类的工作。在分析过程中,分别选取单项、多项能耗指标对抽油机井进行自动分类,以高效地寻找高能耗油井,进而有针对性地对这些油井进行措施调整。

With the help of intellectual oil field infrastructure, production data are now collected automatically in the majority of oil fields. However, the data with large volume and high dimension have not been effectively utilized. In this paper, K-means clustering is applied to analyze energy consumption data for the rod pumping system. The tasks include abnormal data identification and well clustering. During the analyzing process, both single and multiple energy consumption indicators are selected for clustering. In this way, wells with huge energy consumption are found efficiently. This study builds the foundation for well energy consumption saving.