基于小波神经网络的地面集输管道结垢预测研究

Scale Prediction Research on Surface Gathering Pipeline Based on Wavelet Neural Network

小波神经网络是将小波分析和神经网络理论结合起来的一种神经网络,它克服了传统BP神经网络收敛速度慢、学习效率低的缺陷,可以更快、更准确的对一些非线性问题进行研究。文章介绍了小波神经网络的结构、计算与预测的过程。将小波神经网络技术应用于某油田地面集输系统结垢程度的预测研究,其预测结果与实测值相当接近,取得了良好的效果。

The wavelet neural network is one kind of neural networks which closely combines the wavelet analysis with the neural network theory. It can solve the intrinsical limitation of the conventional BP neural network, such as the low convergence speed and learning rate, so as to study nonlinear problems much fast and more accurately. This paper introduces the structure, calculation and prediction process of the wavelet neural network. The wavelet neural network technique is applied to the scale prediction research on the surface gathering pipeline of one oilfield, and the results quite approximate the practical ones and show perfect effects.