DL-based Online Modeling of Nonlinear ODE-beam Cascaded Systems: Application to Flexible Morphing Aircraft

A deep learning (DL)-based online modeling scheme is proposed integrated deep auto-encoder and sparse identification of nonlinear dynamical systems for ordinary differential equation (ODE)-Euler-Bernoulli beam (EBB) cascaded systems. The online model dynamically evolves through new data streams and revising the reconstruction reduced-order dynamic system and reconstruction output for ODE-EBB cascaded systems. Simulation results on a flexible morphing aircraft are provided for validating the effectiveness of the DL-based online modeling approach and its superiority in modeling accuracy and efficiency is also validated by comparing with the DL-based offline one.