Roshani G H, Karami A, Salehizadeh A, Nazemi E
Electrial Engineering Department, Kermanshah University of Technology, Kermanshah, Iran.
Young Researchers and Elite Club, Kermanshah Branch, Islamic Azad University, Kermanshah, Iran.
Appl Radiat Isot. 2017 Nov;129:156-162. doi: 10.1016/j.apradiso.2017.08.027. Epub 2017 Aug 24.
The problem of how to precisely measure the volume fractions of oil-gas-water mixtures in a pipeline remains as one of the main challenges in the petroleum industry. This paper reports the capability of Radial Basis Function (RBF) in forecasting the volume fractions in a gas-oil-water multiphase system. Indeed, in the present research, the volume fractions in the annular three-phase flow are measured based on a dual energy metering system including the Eu and Cs and one NaI detector, and then modeled by a RBF model. Since the summation of volume fractions are constant (equal to 100%), therefore it is enough for the RBF model to forecast only two volume fractions. In this investigation, three RBF models are employed. The first model is used to forecast the oil and water volume fractions. The next one is utilized to forecast the water and gas volume fractions, and the last one to forecast the gas and oil volume fractions. In the next stage, the numerical data obtained from MCNP-X code must be introduced to the RBF models. Then, the average errors of these three models are calculated and compared. The model which has the least error is picked up as the best predictive model. Based on the results, the best RBF model, forecasts the oil and water volume fractions with the mean relative error of less than 0.5%, which indicates that the RBF model introduced in this study ensures an effective enough mechanism to forecast the results.
如何精确测量管道中油气水混合物的体积分数问题仍然是石油工业中的主要挑战之一。本文报道了径向基函数（RBF）预测气-油-水多相系统中体积分数的能力。实际上，在本研究中，基于包括铕和铯以及一个碘化钠探测器的双能计量系统测量环状三相流中的体积分数，然后用RBF模型进行建模。由于体积分数之和是恒定的（等于100%），因此RBF模型只需预测两个体积分数就足够了。在本研究中，采用了三个RBF模型。第一个模型用于预测油和水的体积分数。下一个模型用于预测水和气体的体积分数，最后一个模型用于预测气体和油的体积分数。在下一阶段，必须将从MCNP-X代码获得的数值数据引入到RBF模型中。然后，计算并比较这三个模型的平均误差。选择误差最小的模型作为最佳预测模型。基于结果，最佳RBF模型预测油和水的体积分数，平均相对误差小于0.5%，这表明本研究中引入的RBF模型确保了足够有效的机制来预测结果。
