Abstract

An effective soil classification method is essential for soil layering interpretation. Traditional methods rely on empirical formulae or a general unified system, which may not be accurate for every site. Recent advancements in deep learning have shown promising results in a wide range of practical domains, which provides motivation to explore its potential application in soil classification. A “U-shaped” convolutional neural network (U-Net), coupled with multiple source data to incorporate prior knowledge, is adopted in this study to predict the depths of soil stratum boundaries based on cone penetration testing (CPT), standard penetration testing (SPT) and laboratory index testing data. The U-Net model is first pre-trained on open-access data from other sites around the world and then trained through transfer learning on datasets gathered specifically for the Suzhou No.6 metroline project. Comparison is made from different metrics to investigate how pre-training benefits model performance. The soil boundaries in nine selected output images are extracted and reversed to their original depth for visualization of a subsurface profile. Comparisons to the benchmark interpretations show that the predicted soil profiles have reasonably good agreement with the benchmark profiles developed by an engineering expert. The results indicate that the proposed method is effective and efficient for the prediction of soil stratification.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.