关键词:
Well logging curve stratigraphic comparison
Semantic segmentation
Label smoothing
Attention mechanism
摘要:
Well logging curves serve as indicators of strata attribute changes and are frequently utilized for stratigraphic analysis and *** learning,known for its robust feature extraction capabilities,has seen continuous adoption by scholars in the realm of well logging stratigraphic correlation ***,current deep learning algorithms often struggle to accurately capture feature changes occurring at layer boundaries within the ***,when faced with data imbalance issues,neural networks encounter challenges in accurately modeling the one-hot encoded curve stratifi cation positions,resulting in signifi cant deviations between predicted and actual stratifi cation *** these challenges,this study proposes a novel well logging curve stratigraphic comparison algorithm based on uniformly distributed soft *** the training phase,a label smoothing loss function is introduced to comprehensively account for the substantial loss stemming from data imbalance and to consider the similarity between diff erent layer ***,spatial attention and channel attention mechanisms are incorporated into the shallow and deep encoder stages of U²-Net,respectively,to better focus on changes in stratifi cation *** the prediction phase,an optimized confi dence threshold algorithm is proposed to constrain stratifi cation results and solve the problem of reduced prediction accuracy because of occasional layer *** proposed method is applied to real-world well logging data in oil fi *** evaluation results demonstrate that within error ranges of 1,2,and 3 m,the accuracy of well logging curve stratigraphic division reaches 87.27%,92.68%,and 95.08%,respectively,thus validating the eff ectiveness of the algorithm presented in this paper.