Machine learning, numerical, and analytical approaches for vibration prediction of porous gradient piezoelectric beams under traveling force

[Display omitted] •Application of numerical method, analytical approach, and DT based regression models.•Vibration prediction of porous gradient piezoelectric beams under traveling force.•Rotary inertia, edge conditions, four-parameter foundation, environmental effects. The vibration behavior of tridirectional porous functional gradient (TDPFG) beams coupled with piezoelectric layers under a traveling force is simulated. Moreover, the impressions of force speed and acceleration, four-parameter foundations, and thermo-hygro-magnetic loadings on vibration features are scrutinized. The impacts of different porosity distribution profiles, rotary inertia, and different edge conditions are deliberated in the modeling. Numerical and analytical methods and decision tree (DT) based regression models are exploited for vibration prediction. Transient and steady-state dynamic responses of TDPFG beams and dynamic phenomena are scrutinized. Various comparative and parametric investigations are conducted to confirm and appraise the current results. The effectiveness and efficiency of introduced machine learning (ML) methodologies are examined. It is recognized that the fine DT classifier is the most successful model for predicting the TDPFG beam vibration. Besides, it is found that compared with numerical and analytical treatments, the proposed DT-based algorithms offer lower and higher computational time to determine the vibration features of TDPFG beams with suitable performance (high determination coefficient value as well as low mean absolute error and mean square error values). It is also understood that optimal material attributes can cancel undesirable system vibration. The presented results are advantageous in designing and optimizing state-of-the-art high-speed transportation systems.