Stratigraphic automatic correlation using SegNet semantic segmentation model

Dai, Yue (Southwest Petroleum University) | Huang, Xuri (Southwest Petroleum University) | Liu, Haojie (Sinopec Shengli Oilfield Company) | Yang, Hongwei (Sinopec Shengli Oilfield Company) | Wei, Guohua (Sinopec Shengli Oilfield Company) | Lu, Ning (Sinopec Shengli Oilfield Company) | Han, Zhiying (Sinopec Shengli Oilfield Company) | Song, Haibo (Beijing Sunrise PetroSolutions Tech., Ltd.)

OnePetro 

The second is that automatic stratigraphic correlation algorithms are difficult to achieve highresolution A multi-task encoder-decoder based on SegNet architecture stratigraphic correlation based on multi-source is proposed for automatic stratigraphic correlation in this data. Moreover, the traditional methods can not have work. In order to have higher resolution correlation, logs automatic multi-well processing, which leads to time-and and their wavelet transformed results are combined to form labor-consuming. the training datasets. In addition, two types of loss functions for the SegNet are used to achieve high-resolution results. In recent years, the emerging of artificial intelligence has By applying this method to a field in Shengli Oilfield, the inspired deep learning technique application for automatic result demonstrates that this network can obtain accurate stratigraphic correlation. Wu and Nyland (1987) established stratigraphic correlation and is significantly efficient an artificial intelligence stratigraphic interpretation system compared to the conventional manual method. Using the based on linear regression model and heuristic algorithm.

Duplicate Docs Excel Report

Title
None found

Similar Docs  Excel Report  more

TitleSimilaritySource
None found