Git Product home page Git Product logo

shagunsharma98 / global-structural-earthquake-damage-prediction Goto Github PK

View Code? Open in Web Editor NEW
13.0 1.0 2.0 725 KB

Using deep learning techniques like 1D and 2D CNNs, LSTM to detect damage in a structure with hinges/joints after an earthquake.

Jupyter Notebook 100.00%
earthquakes damage-detection convolutional-neural-networks lstm-neural-networks spectrogram structural-engineering grid-search-hyperparameters training-1d-cnn 2d-cnns spectrograms

global-structural-earthquake-damage-prediction's Introduction

Global-Structural-Earthquake-Damage-Prediction

Using deep learning techniques to detect damage in a structure(G+20 building) with hinges formation in joints after an earthquake. The networks used to train the data includes: '1D Convolutional Neural Network', 'LSTM Network', '2D Convolutional Neural Network using spectrograms'. Data is recorded for X & Y axes using accelerometers attached to the joints of the structure and the cases of earthquakes are generated by providing impulses to the structure using SAP2000.

Data for training can be found at (Please refer to the .zip file of the dataset):

https://drive.google.com/file/d/1zBAMNiHTHqJS5fIZSiklUlsfCli9HM82/view?usp=sharing

For data used in the Predictions notebook, refer to:

https://drive.google.com/drive/folders/1Gzo2Zqh0XB43IqyGRfU_rCcj8ABOBTeY?usp=sharing

Once you check the data, file 'Data and Discrepancies.txt' describes the discrepancy found in each of the cases in data.

  • Refer to file 'Earthquake Damage Detection-2 Data Preprocessing.ipynb' for processing the data.

  • Refer to the file 'Earthquake Damage Prediction-2 Preparing I&O.ipynb' for preparing the data to be fed into networks and generating the spectrograms for both X&Y axes.

  • Refer to the file 'Earthquake Network Training 3.ipynb' for training 1D CNN and LSTM. The attached notebook might contain the code for training the 2D CNN using spectrograms as well but the training is done in a seperate notebook.

  • Refer to the file 'Earthquake Network Training 2 Layered 1D CNN & Spectrogram Training.ipynb' for training a 2 Layered 1D CNN (as a part of multiple experimentations performed using 2, 3, 4 layers of 1D CNNs) and the 2D CNNs using spectrograms.

  • Refer to the notebook 'Copy of SHM_Notebook.ipynb' for Grid Searching the optimized hyperparameters in 1-D CNN and LSTM (since both outperformed the conventional 2-D CNN and would later be ensembled in future developments to improve the predictions).

  • Refer to the notebook 'Predictions.ipynb' to refer to the predictions provided by the trained networks on an untouched data.

global-structural-earthquake-damage-prediction's People

Contributors

shagunsharma98 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.