Predicting CBR values using gaussian process regression and meta-heuristic algorithms in geotechnical engineering

The California Bearing Ratio (CBR) test is a vital tool in geotechnical engineering and is useful in controlled laboratory settings and dynamic field applications. It plays a pivotal role in determining the load-bearing properties of subgrade soil, which is essential for various construction project...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Multiscale and Multidisciplinary Modeling, Experiments and Design Ročník 7; číslo 4; s. 3799 - 3813
Hlavní autoři: Wu, Xu, Yang, Feng, Huang, Shuchen
Médium: Journal Article
Jazyk:angličtina
Vydáno: Cham Springer International Publishing 01.09.2024
Témata:
Characterization and Evaluation of Materials
Engineering
Mathematical Applications in the Physical Sciences
Mechanical Engineering
Numerical and Computational Physics
Original Paper
Simulation
Solid Mechanics
Gaussian process regression
Leader Harris Hawk’s optimization
California bearing ratio
Dynamic arithmetic optimization algorithm
ISSN:2520-8160, 2520-8179
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The California Bearing Ratio (CBR) test is a vital tool in geotechnical engineering and is useful in controlled laboratory settings and dynamic field applications. It plays a pivotal role in determining the load-bearing properties of subgrade soil, which is essential for various construction projects, including retaining walls, highway embankments, bridge abutments, and earth dams. CBR values obtained from this test are fundamental for assessing soil strength and integrity, making it a cornerstone of geotechnical engineering. This paper presents an innovative method for predicting CBR values with high precision. It employs Gaussian Process Regression (GPR) to develop complex and highly accurate predictive models. These models encompass a wide range of soil characteristics, such as particle distribution, plasticity, linear shrinkage, and the type and quantity of stabilizing additives. The GPR model significantly improves predictive modeling accuracy by establishing robust relationships between these soil attributes and CBR values. Additionally, the study incorporates two advanced meta-heuristic algorithms, the Dynamic Arithmetic Optimization Algorithm (DAOA) and Leader Harris Hawk’s Optimization (LHHO), to enhance the precision of the predictive model. This collaborative effort resulted in the creation of three models: GPR + LHHO (GPLH), GPR + DAOA (GPDA), and GPR. The GPDA model stands out with exceptional predictive capabilities, achieving a remarkable R 2 value of 0.989 during training and an optimal RMSE of 1.488, confirming its precision and consistency. This innovative approach advances CBR prediction and reinforces the reliability of geotechnical engineering practices across diverse soil conditions, making it a significant contribution to the field.
ISSN:2520-8160
2520-8179
DOI:10.1007/s41939-024-00428-0