【摘要】 针对油田生产过程中的在线多相流实时计量这一难题,本文提出了一种基于多传感器和SVR算法的油田油气水多相流量量化检测方法,采用先进的文丘里和微波检测手段,对气相、液相流量和液相含水率等多相流关键参数进行测量,并结合基于支持向量回归的机器学习算法进行测量模型构建和评价,评价结果表明:本文提出的多相流测量模型的测量精度可以达到气相、液相流量和液相含水率计量误差<10%的工业多相流计量要求,本文的研究成果为油气生产过程中多相流在线实时测量提供了一项新的技术手段。更多还原

【Abstract】 Aiming at the problem of real-time measurement of on-line multiphase flow in oilfield production process, this paper presents a quantitative multiphase flow measurement method for oil, gas and water based on Multi-sensor and SVR algorithm. The key parameters of multiphase flow, such as gas phase, liquid phase flow and liquid phase water content(wlr), are measured by advanced venturi and microwave measurement methods, and combined with the machine learning algorithm based on support vector regression(SVR) to construct and evaluate the measurement model. The evaluation results show that the measurement accuracy of the multi-phase flow measurement model proposed in this paper can meet the requirements of industrial multi-phase flow measurement, with measurement errors of gas phase, liquid phase flow and liquid phase water content less than 10%. This study provides a new technical for on-line real-time measurement in oil-gas production process.更多还原