This repository presents data and code of our published paper.
Jian X, Xia Y, Chatzi E, Lai Z. Bridge influence surface identification using a deep multilayer perceptron and computer vision techniques. Structural Health Monitoring. DOI: 10.1177/14759217231190543
The 'Train_MLP.py' is a Python script used to train an MLP to identify bridge influence surfaces from simulated vehicle-bridge interaction data (wheel_loc.csv and u_raw.csv).
The 'Evaluate_MLP.m' is a MATLAB script used to evaluate the accuracy of bridge influence surfaces identified by MLP.
The 'Evaluate_Quilligan.m' is a MATLAB script used to evaluate the accuracy of bridge influence surfaces identified by the widely-accepted Quilligan's method.
Please contact xudong.jian@sec.ethz.ch should you have any questions.