Enhancing Undrained Shear Strength Prediction through Innovative Hybridization Techniques
Ensuring precise evaluation of undrained shear strength (USS) in soft, sensitive clays is critical in geotechnical engineering. precision in predicting USS holds utmost significance in safeguarding the structural soundness and steadiness of foundations and earthworks. The issue is addressed by emplo...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Bilijipub publisher
2024-03-01
|
| Series: | Journal of Artificial Intelligence and System Modelling |
| Subjects: | |
| Online Access: | https://jaism.bilijipub.com/article_193313_014b80aded1e0a72b5b21b3332f869d4.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Ensuring precise evaluation of undrained shear strength (USS) in soft, sensitive clays is critical in geotechnical engineering. precision in predicting USS holds utmost significance in safeguarding the structural soundness and steadiness of foundations and earthworks. The issue is addressed by employing cutting-edge data-driven Decision Tree (DT) techniques to analyze the intricate connections between USS and fundamental soil parameters. An innovative approach for predicting USS in soft clays using machine learning methods is pioneered in this study, with a specific focus on a set of five crucial feature variables, such as pre-consolidation stress (PS), vertical effective stress (VES), liquid limit (LL), plastic limit (PL), and natural water content (W), which serve as inputs for the data-driven models. The selection of these variables is based on their established impact on USS. Notably, this study adopts an innovative approach by employing the Alibaba and the Forty Thieves (AFT) and Aquila Optimizer (AO) methods to refine the model's hyperparameters, leading to a reduced reliance on heuristic methods and computationally expensive brute-force searches. This streamlined methodology enhances the accuracy of USS predictions and optimizes the model's efficiency. As a result, DTAO obtained a more suitable performance compared to other models, with R2 and RMSE equal to 0.994 and 76.142, respectively. The research outcomes have the potential to provide substantial advantages to geotechnical engineers and researchers. They offer a robust, data-driven approach for evaluating USS in soft, sensitive clays, thereby contributing to the advancement of safety and stability in civil engineering projects. |
|---|---|
| ISSN: | 3041-850X |