This project uses a Physics-Informed Neural Network (PiNN) to analyze slope stability. It combines physics with neural networks to predict slope deformation and failure.
The PiNN models a 2D slope problem using physics equations (linear elastic and gravity). It aims to predict slope stability in a way similar to traditional methods like Finite Element Method (FEM).
- Predicts slope stability using PiNN.
- Compares PiNN results with FEM.
- Shows displacement and deformation.
- Handles gravity loading.
- Exports displacement data for plotting.
The PiNN is trained using:
- Physics equations: 2D linear elastic
- Material parameters:
- Young's Modulus (E): 50000 kN/m^2 (for example) to compare with FEM from Plaxis2d
- Poisson's Ratio (ν): 0.3
- Unit Weight (γ): 18 kN/m³ (for example) to compare with FEM from Plaxis2d
- Boundary Conditions: Fixed displacements on the bottom, left, and right sides.
- Raissi, M., Perdikaris, P., & Karniadakis, G. E. (2019). Physics-Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Differential Equations. Journal of Computational Physics, 378, 686–707.
- Apisit Robjanghvad : M.eng (Geotechnical engineering student), Department of Civil Engineering King Mongkut's University of Technology Thonburi (KMUTT) Email: [apisit65a@gmail.com]
Install the necessary tools using:
pip install torch matplotlib numpy
pip install pandas
pip install pytorch