Full text loading...
-
High-Resolution Reservoir Prediction Using Convolutional Neural Networks
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 81st EAGE Conference and Exhibition 2019, Jun 2019, Volume 2019, p.1 - 5
Abstract
We propose a new method to predict the reservoir distribution based on convolutional neural networks by using a few well log data. Deep learning methods usually need a large amount of samples, while the well log data is usually limited. To train the convolutional neural network properly by the limited log well data, we propose a two-step strategy. In the first step, the network is pre-trained by a rough prediction result which is obtained by the traditional method. Then, in the second step, the pre-trained network is fine-tuned by well log data which can be regarded as accurate labels. Field data examples demonstrates that the deep network trained based on this strategy can achieve more accurate with a high resolution reservoir prediction results.