| With the increasing length of high-speed railway construction lines in China,the speed of train operation is faster and faster.Environmental vibration caused by running trains has a negative impact on residents,precision instruments and buildings along the line.Environmental vibration caused by rail transit has attracted more attention from researchers at home and abroad.In this paper,based on the combination of numerical simulation and artificial intelligence algorithm,the vibration prediction of surrounding soil and buildings caused by elevated train operation is studied.The main researchcontents are as follows:(1)A three-dimensional numerical analysis model of train-track-bridge-soil-building is established.The model includes two subsystem models:train-track and bridge-soil-building.The former model calculates the off-track force as the excitation load of the latter model.In the model,the influence of dynamic interaction between pile foundation and soil,soil and building foundation is considered.(2)By changing the input parameters of the numerical model,the changes of train speed,marshalling,distance,properties of foundation soil and height of buildings are analyzed,and the influence degree of vibration response of surrounding soil and buildings caused by elevated train is summarized.(3)Based on the training of numerical simulation data,the BP neural network prediction model is established,and on the basis of the measured data,the prediction model is self-learning and retraining,so as to improve the prediction accuracy of environmental vibration caused by elevated train.The validity and feasibility of the prediction model are verified by comparing the prediction data of environmental vibration caused by elevated train with the numerical simulation results and the measured data.In this paper,the artificial intelligence algorithm is applied to the prediction of environmental vibration caused by elevated trains,which provides a fast method for predicting environmental vibration caused by high-speed trains,and also provides a new idea for the study of similar problems. |
