原油管道内蜡沉积速率预测及分析

分别运用灰色预测理论和灰色神经网络理论对原油管道内的蜡沉积速率进行了预测分析；应用灰色人工神经网络理论，考虑剪切应力、温度梯度、粘度以及浓度梯度4个影响因素作为主要因素的对原油管道内的蜡沉积速率进行的预测，与传统的灰色预测方法相比，所得到的预测值更为接近实际值，蜡沉积速率的相对误差绝对值在1．6％以内，灰色神经网络用于管道内蜡沉积速率预测的效果良好，能为原油管道蜡沉积规律的深入研究和制定合理的清蜡周期提供理论依据。

Prediction and Analysis on Wax Deposition Rate in the Crude Oil Pipeline

The wax deposition rate in the crude oil pipeline was predicted and analyzed with grey theory and grey neural network theory, respectively. Based on considering the shear stress, temperature, viscosity and concentration gradients as main factors, the wax deposition rate in the crude oil pipeline was predicted with the theory of grey and artificial neural network. The results show that, compared with the traditional grey theory, the prediction is more close to the actual value, the relative error is less than 1.6%, the effect of predicting wax deposition rate with the theory of the grey neural network is good, it can provide a theoretical basis of the further study on the regular pattern of the wax deposition rate and formulating a reasonable period of removing the wax for the crude oil pipeline.