基于人工神经网络的含水原油视粘度计算
Calculation on the Apparent Viscosity of Watered Oil Based on Artificial Neural Network
-
摘要: 对大庆含水原油视粘度与含水率、剪切速率和油温的关系进行了全面研究, 结果表明, 含水原油的视粘度是温度、含水率以及剪切速率的函数。在转相点以前, 视粘度随含水量的上升而增加, 且受温度影响较大, 同时剪切速率的影响也相当明显。随着剪切速率的增加, 转相点的视粘度明显下降。在转相点以后, 视粘度随含水量的增加而降低, 且受温度和剪切速率影响, 乳状液视粘度进入高含水区后变化趋于平缓。运用人工神经网络的方法来模拟各种影响因素与含水原油视粘度之间的映射关系, 建立了含水原油视粘度计算公式。这种方法综合考虑了温度、含水率以及剪切速率对含水原油的视粘度的影响, 模型计算精度高, 为准确计算高含水原油管道工艺参数奠定了基础。Abstract: The relationship between the apparent viscosity of watered oil and its water cut, temperature and shear rate is studied and the oil in Daqing Oilfield is taken as an example. The result shows that the apparent viscosity of watered oil is the function of the temperature, water cut and shear rate. Before the phase transition point, the apparent viscosity of oil increases with the increase of water cut, and temperature and shear rate have great effects on it. The apparent viscosity at the point decreases distinctly with the increment of shear rate. After the phase transition point, the apparent viscosity of it decreases with the increment of water cut, and temperature and shear rate have also effects on it. Using neural network to simulate the relationship between the apparent viscosity of watered oil and its effect factors, the formula of calculation the apparent viscosity of watered oil is established. With this way, temperature, shear rate and water cut are taken into account and the model is very precision. The study of this paper establishes a foundation for calculating precisely pipeline technological parameter.