Architectural Design and Active Vibration Suppression of Sandwich Plates Using Deep Neural Network-Based Validation

This research explores the architectural design and active vibration suppression of three-layer rectangular sandwich plates through a deep neural network (DNN)-based verification framework. The proposed structure has been designed as a sandwich plate subjected to a time-dependent external force and consists of the piezoelectric actuator and sensor face sheets bonded to a zinc oxide–graphene oxide (ZnO–GO) hybrid nanocomposite-reinforced polymer core. The structural mechanics are modeled by the Carrera unified formulation (CUF), allowing a systematic hierarchical representation of displacement fields. The unified formulation produces the governing equations for dynamic response and control via an appropriate discretization strategy and a mixed-interpolation finite element technique. These equations are switched to the Laplace domain for computational efficiency, and the corresponding time-domain solutions are retrieved through the modified Dubner–Abate (MDA) inversion method. A wide range of control schemes has been applied to achieve active vibration suppression, including simple linear damping (SLD) controllers, adaptive band-limited derivative (ABL-D) controllers, hysteresis-based nonlinear (HBN) controllers, fuzzy logic supervisory (FLS) controllers, and hybrid predictive sliding mode (HPSM) controllers. Comparative studies show that the HPSM controller is more robust, has better performance in terms of vibration reduction, and is more stable under different loading and material conditions. In order to ensure the reliability of the control strategies in a more certain way, a DNN-based verification framework is developed to compare the predicted vibration responses with the finite-element-derived control outputs. The outcomes from the deep neural networks (DNNs) show a significant relationship with the predicted structural responses that, in turn, support the credibility of the CUF method and the management system.