TL-RNN-StatePrediction-StructDyn

This repository contains the code for the paper "State Estimation in Structural Dynamics through RNN Transfer Learning". Feel free to contact us for any questions or comments.

Installation

Clone the repository and navigate to the directory.

git clone https://github.com/shuohaopolyu/TL-RNN-StatePrediction-StructDyn.git
cd TL-RNN-StatePrediction-StructDyn

Install the conda package manager from here.

conda create --name tlrnn python=3.10.14
conda activate tlrnn

Install the required packages.

conda install --yes --file requirements.txt

Usage

The code is organized in the following way:

• main.py: Contains the main program for running the experiments.
• models/: Contains the RNN models used in the paper.
• experiments/: Contains experiments for the state estimation.
• systems/: Contains finite element models of the structures and solvers for response simulation.
• excitations/: Contains the excitation models for the finite element simulations.
• dataset/: Contains the data used in the paper.
• figures/: Contains the program for generating the figures in the paper.
• utils.py: Contains utility functions for data analysis, processing, and generation.

To run the experiments, use the following command:

python main.py

Note: Depending on which steps of the experiment you wish to run, you may need to uncomment or comment certain lines in main.py to include or exclude specific functions. This allows you to execute different parts of the experiment as needed.

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.