前馈神经网络用于压裂效果预测

Prediction of fracturing effect using feed-forward neural network

江汉测井研究所,潜江433123

Jianghan Well Logging Institute, Qianjiang 433123

本文提出用前馈神经网络进行油田压裂效果预测,并介绍了几种行之有效的加速学习算法。

This paper presents an approach, based on feed forward neural network, to predict formation fracturing effects and adopts several methods to increase the learning rate of EBP algorithm. The results show that feed forward neural network algorithm can better and efficiently obtain the prediction of formation fracturing effects.