基于数据融合技术的水下油气井计量方法研究

Flow Metering Method for Oil and Gas Wells Based on Data Fusion Technique

虚拟流量计系统具有成本低、精度高等优点,逐渐成为水下油气井生产计量的一种新方式,其核心是依托多相流模拟技术的油气井流量算法。为了提高虚拟流量计系统在线应用时的可靠性和精度,基于数据融合技术和多个物理流量模型,设计了分别基于D-S证据理论、不确定度理论和神经网络理论的3种融合流量估计方法。现场生产数据测试结果表明,数据融合流量估计方法的平均误差维持在5%以内,且当单个物理流量模型失效时,融合技术仍能给出最优的计量结果,保证了整个虚拟流量计系统的可靠性。

Virtual flowmeter system based on oil and gas well flow algorithm of the multi-phase flow simulation technology,has become a new measurement method for underwater oil and gas well production by its low costs and high precision. In order to improve the reliability and accuracy of this online system,three methods of fusion flow evaluation based on the theory of D-S evidence,uncertainty and neutral network were designed from the data fusion technology and multiple physical flow models. The verification results of field production data show that the average error of fusion flow estimation methods could be kept within 5% and the fusion flow estimation method can provide the optimal measurement results even at single physical flow model failure.