Nonlinear and Distributed Control of Smart Structures Using Artificial Neural Networks
Pagination or Media Count:
This interdisciplinary research program is focused on designing and implementing robust controllers on smart structures. The primary research objectives are 1 structural modeling of smart structures incorporating both material and geometric nonlinearities, 2 development of controllers using conventional and neural network based algorithms, and 3 distributed control techniques using spatially distributed sensors and actuators. We have investigated modeling techniques for structures that account for the material and geometric nonlinearities, continuity of tractions across interlaminates, hysteric behavior of the FZTs, interaction between the PZTs and the substrate, orthotropic nature of the PZTs, anisotropic properties of composite plates, and the inertia forces. A structural identification method based on the measurement of eigenvalues and eigenvectors of the structure has been developed and verified using experimental results. We have successfully designed and implemented multiobjective robust controllers as smart structures. Practical considerations and limitations of the controller implementation using neural network hardware are also investigated.