Adaptive Active Vibration Control of Sandwich Smart Plates with Environmentally Friendly Coarse Aggregate Ultra-High Performance Concrete Core: Application of Physics-Informed Neural Networks Solution Procedure and Carrera Unified Formulations

This study presents an advanced approach for the adaptive active vibration control of sandwich smart plates with environmentally friendly coarse aggregate ultra-high performance concrete (UHPC) cores, incorporating sensor and actuator face sheets. A physics-informed neural networks (PINNs) solution procedure is applied in conjunction with the Carrera unified formulations (CUF) to effectively model and analyze the complex structural dynamics of such smart composite systems. The environmentally conscious selection of UHPC, enhanced with coarse aggregates, ensures sustainability while maintaining superior mechanical performance. The adaptive control system is developed using embedded piezoelectric sensors and actuators, which dynamically regulate structural vibrations under varying operational conditions. The proposed computational strategy integrates PINNs to solve the governing partial differential equations while enforcing boundary and initial conditions through neural network training, thus reducing dependence on traditional meshing techniques. To ensure the robustness of the PINNs-based approach, deep neural networks (DNNs) are presented for verification of the mathematical simulation results, confirming the fidelity of the predicted displacement and stress fields. This dual neural network framework significantly enhances simulation efficiency and predictive accuracy. This study demonstrates that coupling CUF with PINNs provides a flexible and scalable framework for smart structure analysis, contributing to the development of sustainable, intelligent materials for next-generation civil and aerospace engineering applications.