A Comparative Study of Hybrid Adaptive Neuro-Fuzzy Inference Systems to Predict the Unconfined Compressive Strength of Rocks
The exact determination of Unconfined Compressive Strength (UCS) in rock samples is essential for mining and civil engineering projects to be planned and carried out efficiently. However, the inherent variability and discontinuity within rock masses present challenges in obtaining accurate physico-m...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Bilijipub publisher
2024-06-01
|
| Series: | Advances in Engineering and Intelligence Systems |
| Subjects: | |
| Online Access: | https://aeis.bilijipub.com/article_199131_6501fff2918d8b546e6045e46647a480.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The exact determination of Unconfined Compressive Strength (UCS) in rock samples is essential for mining and civil engineering projects to be planned and carried out efficiently. However, the inherent variability and discontinuity within rock masses present challenges in obtaining accurate physico-mechanical data. In this study, the application of Adaptive Neuro-Fuzzy Inference Systems (ANFIS) is investigated as a cost-efficient and reliable approach to predict UCS from Bulk Density (BD), Bulk Tensile Strength (BTS), dry density (DD) test, p-wave velocity test ( ), Schmidt hammer rebound number (SRn), and point load index test (Is(50)). ANFIS excels in effortlessly amalgamating expert knowledge using linguistic rules and empirical data, thus enhancing model interpretability and transparency, qualities profoundly esteemed in the geosciences. Furthermore, it allows for parameter fine-tuning optimizing model performance, as evidenced by its enhanced performance with the Mountain Gazelle Optimizer (MGO) and Adaptive Opposition Slime Mold Algorithm (AOSMA) in this study. Models were trained with 70% of the data, and the performance of the models was tested with 30% of the dataset. Performance metrics like R2, RMSE, NMSE, MAE, and n_10 index were used to assess the predictive capability of models, indicating that ANAS with maximum and minimum =3.103, has the most optimal prediction performance for practical applications. |
|---|---|
| ISSN: | 2821-0263 |