Structural Damage Detection using a hybrid CNN-SVM

This repository contains a CNN-SVM for structural damage detection using acceleration data. The model is implemented in Python using the Pytorch library. Note: This is an element-by-element monitoring of the structure.

Requirements

To use this code, you will need the following libraries:

• torch
• matplotlib
• numPy
• sklearn
• pandas

You will also need access to the appropriate data files for training and evaluating the model.

Data Loading

The data_reader function is responsible for loading the data that will be used to train and test the model. It has the following parameters:

• CSVpath: a string containing the file path to the Excel file containing the addresses of the data.
• col: a list of integers specifying the columns of the data to use.
• skiprows: an integer specifying the number of rows to skip at the beginning of the data file.
• max_rows_undamaged: an integer specifying the maximum number of rows to load from the undamaged data files.
• max_rows_damaged: an integer specifying the maximum number of rows to load from the damaged data files.
• resample_factor: an integer specifying the resampling factor for the data. If set to 1, the data will be downsampled by a factor of 2 using odd rows. If set to 2, the data will be upsampled by a factor of 2 using the even rows.
• batch: batch size

Training and Evaluating the Model

The model is fit to the training data. The model is then used to make predictions on the test and the training data. The performance of the model is evaluated based on accuracy, f1 score, recall, precision, and roc_auc_score.

Additional Resources

In case you found this architecture useful for your research, please consider citing our book chapter as follows:

Ghazvineh S, Nouri G, Hosseini Lavasani S, Gharehbaghi V, Noroozinejad Farsangi E, Noori M. 7 Vibration-based damage detection using a novel hybrid CNN-SVM approach. In: Noori M, Yuan F, Farsangi E (ed.) Data-Centric Structural Health Monitoring: Mechanical, Aerospace and Complex Infrastructure Systems. Berlin, Boston: De Gruyter; 2023. p.137-158. https://doi.org/10.1515/9783110791426-007

If you have any questions or need further assistance, please open an issue in this repository.