Development of automatic prediction model for ground vibration using support vector machine

This study develops an automatic prediction model for ground vibration induced by Taiwan high-speed trains on embankments. Various field-measured ground vibration data comprise a database used for developing the prediction model. First, the main characteristics that affect the overall vibration level are established on the basis of measurement result database. These main influence factors include train speed, ground condition, measurement distance, and supported structure. A support vector machine (SVM) algorithm, which is a widely used classification model, is then adopted to predict the vibration level induced by high-speed trains on embankments. The measured and predicted vibration levels are compared to verify prediction model reliability. Analytical results show that the more the measured vibration data located in one vibration group is collected, the higher of the accuracy rate will be. Generally, the developed SVM model can reasonably predict ground vibration level in various frequency ranges. Prediction results are discussed in detail, and the methodology for developing the automatic ground vibration level prediction system is presented.