Scholarlink

In order to accurately and efficiently predict the building vibration level caused by underground train operation, four algorithms, namely RF, SVM, XGBoost, and LightGBM, are employed in this paper. These algorithms have been trained on the basis of a database generated from numerical simulations. Firstly, this paper sought to elucidate the relationship between the five variables. Subsequently, four metrics (R2, MSE, MAE, and RMSE) were employed to evaluate the performance of the four machine learning algorithms. Moreover, the correlation coefficient matrix heat map demonstrates that the intensity of vibration is significantly correlated with the speed of metro train operation, horizontal distance, and number of floors. A comparison of the four metrics indicates that the XGBoost model is the most effective in predicting vibration levels.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intelligent Prediction of Vibration Caused by Metro Train Operation Based on Machine Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shulin Zhu,
  • Xiaolei Zhang,
  • Jian Wu,
  • Yong Feng

摘要

In order to accurately and efficiently predict the building vibration level caused by underground train operation, four algorithms, namely RF, SVM, XGBoost, and LightGBM, are employed in this paper. These algorithms have been trained on the basis of a database generated from numerical simulations. Firstly, this paper sought to elucidate the relationship between the five variables. Subsequently, four metrics (R2, MSE, MAE, and RMSE) were employed to evaluate the performance of the four machine learning algorithms. Moreover, the correlation coefficient matrix heat map demonstrates that the intensity of vibration is significantly correlated with the speed of metro train operation, horizontal distance, and number of floors. A comparison of the four metrics indicates that the XGBoost model is the most effective in predicting vibration levels.