AI-powered simulation models for estimating the consolidation settlement of shallow foundations

The shallow foundations are one of the commonly used, cost-effective and versatile substructure in the infrastructure and geotechnical society. The consolidation settlement is one of the influential parameters for the design purpose of the shallow foundation. This study utilized the AI based models like Deep Neural Network (DNN), Random Forest (RF) and Gradient Boosting Machine (GBM) for the prediction of S_c. In order to forecast the S_c, different soil conditions such as void ratio, compression index, density and the load were considered as the input criteria and their respective settlement is the output. These adopted AI driven models, provide better results with higher precisions. The output produced by the adopted models were considered for different statistical assessments, specifically, DNN model exhibits its supremacy in terms of precision and erros (R² = 0.9992 & RMSE 0.6404) than RF (R² = 0.9977 & RMSE = 1.261) and GBM model (R² = 0.9838 & RMSE = 2.911). Moreover, the rank analysis, Taylor diagram and the reliability index were also computed for justifying the capability of the developed AI models. These findings have significant implications for the accurate prediction of consolidation settlement in shallow foundations, enabling more reliable and efficient designs. By providing precise S_c predictions, these models contribute to improved foundation design, reducing the risk of excessive settlement and ensuring structural integrity.