Attenuation characteristics and machine learning predictions for metamaterial foundation systems

The ambient vibrations caused by trains have become more serious. It is imperative to isolate those vibration. Thus, this proposes a novel metamaterial foundation based on local resonance theory for reducing vibration induced by trains and establishes a neural network model. The bandgaps of in-plane vibration and out-plane vibration are 30.1∼52.1Hz and 13.8∼25.5Hz. This indicates that the metamaterial foundation can be used to reduce vibrations induced by trains, and the neural network model has good performance, can accurately predict the bandgap based on structural and material parameters.