Font Size: 

Rapid seismic damage evaluation of subway stations is critical for the efficient decision on the repair methods to damaged subway stations caused by earthquake and rapid recovery of subway networks without much delay. However, the current methods to evaluate the damage state of a subway station after earthquake is mainly field investigation by manual or computer vision, which is dangerous and time-consuming. In view of this, a novel methodology which adopting machine learning technologies as classification model to rapidly and accurately evaluate the damage state of subway stations after earthquakes is proposed in this study. Four commonly used machine learning technologies including Artificial Neural Networks (ANN), Support Vector Machine (SVM), Random Forest (RF) and Logistic Regression (LR) are selected as the possible classification models. A subway station in Shanghai is selected as a case to demonstrate the proposed methodology. To generate samples for training and testing the classification models, the numerical model of the selected subway station is established and validated by a model test. The four classification models are then established using the training set and their performance are evaluated and compared using test set. The results show that the RF provides the best performance for seismic damage evaluation of subway stations. In addition, the most influential intensity index of ground motion is obtained by RF.

machine learning