ISSN 1673-8217 CN 41-1388/TE
主管:中国石油化工集团有限公司 主办:中国石油化工股份有限公司河南油田分公司
高翔, 王云峰, 刘海波. 2020: 基于大数据挖掘技术不停机间抽工作制度优化——以大庆油田为例. 石油地质与工程, 34(05): 109-113.
引用本文: 高翔, 王云峰, 刘海波. 2020: 基于大数据挖掘技术不停机间抽工作制度优化——以大庆油田为例. 石油地质与工程, 34(05): 109-113.
GAO Xiang, WANG Yunfeng, LIU Haibo. 2020: Optimization of non-stop intermittent pumping system based on big data mining technology——by taking Daqing oilfield as an example. Petroleum Geology and Engineering, 34(05): 109-113.
Citation: GAO Xiang, WANG Yunfeng, LIU Haibo. 2020: Optimization of non-stop intermittent pumping system based on big data mining technology——by taking Daqing oilfield as an example. Petroleum Geology and Engineering, 34(05): 109-113.

基于大数据挖掘技术不停机间抽工作制度优化——以大庆油田为例

Optimization of non-stop intermittent pumping system based on big data mining technology——by taking Daqing oilfield as an example

  • 摘要: 针对不停机间抽技术提高抽油机井泵效和降低能耗等方面效果显著,但工作制度确定过程中易出现实际确定方法操作性不强,主观因素影响大,单井设计个性化弱等问题,通过分析不停机间抽工作制度的影响因素,在优选出9个相对独立因素的基础上,利用大数据挖掘技术,以不停机间抽正常运行时间和运行周期为分析挖掘对象,对比分析了常用数据挖掘算法的适应性,并优选出适应性最强的算法。结果表明,回归计算中,BPNN算法要优于R– SVM和MRA算法;分类算法中,C–SVM算法比BAYSD、NBAY等算法更优;使用C–SVM—BPNN算法对不停机间抽工作制度优化后系统效率和泵效明显提高。研究结果对于不停机间抽井确定最优工作制度具有较好的指导作用。

     

    Abstract: The non-stop intermittent pumping wells are good to improve the pumping efficiency and reduce the energy consumption, while in the process of determining the working system, there are many problems, such as the low operability of the actual determination method, the great influence of subjective factors, and the weak individualization of single well design. By analyzing the influencing factors of the working system, based on the optimization of 9 relatively independent factors, by using big data mining technology, taking the normal running time and running period as the analysis mining object, the adaptability of common data mining algorithms is compared and analyzed, and the algorithm with the strongest adaptability is selected. The results show that BPNN algorithm is better than R-SVM and MRA algorithm in regression calculation, C-SVM algorithm is better than BAYSD and NBAY algorithm in classification algorithm, and C-SVM-BPNN algorithm is better than BAYSD and NBAY algorithm in system efficiency and pump efficiency. The research results have a good guiding role in determining the optimal working system of non-stop intermittent pumping wells.

     

/

返回文章
返回