This project explores the application of Hierarchical Bayesian Modeling (HBM) to derive geotechnical parameters, enhancing the reliability of soil strength assessments across multiple sites. By leveraging Bayesian inference, the project aims to quantify uncertainties and improve the robustness of geotechnical analyses.

Project Overview

Accurate estimation of soil strength parameters is crucial for geotechnical engineering applications. This project focuses on:

  • HBM Framework: Implementing a hierarchical Bayesian approach to model variability in soil properties across different locations.
  • Data Integration: Combining site-specific data with global information to improve parameter estimation.
  • Uncertainty Quantification: Assessing the confidence in derived parameters to inform risk-based decision-making.

Technical Highlights

  • Bayesian Modeling: Orchestration of Hierarchical Bayesian models for capturing both the local behaviour and global trend of the data at the same time.
  • Bayesian Inference: Utilizing Hamiltonian Monte Carlo (HMC) methods for inference.
  • Software Tools: Employing a statistical computing environment in PyMC for Bayesian analysis.

Results

The application of HBM provides a more comprehensive understanding of soil strength variability, leading to improved predictions and risk assessments in geotechnical engineering projects.

🔗 View Paper