Optimal design methodology for an active structure using simultaneous optimization
Department of Mechanical Engineering-Engineering Mechanics
In this paper, a new simultaneous optimization methodology is proposed and applied to active control of piezoelectric laminate beams. A finite element method is used to model the optimal distribution of a conventional beam structure, a homogenization approach is developed to allocate piezoelectric materials to obtain smart sensors/actuators, and a linear quadratic regulator (LQR) is applied for the control law. The parallel genetic algorithm based active structure design code was developed, including ABAQUS® for model development and Matlab® for control design. During the optimization process, structural material and smart material were removed from the cells to minimize a cost function which based on the control weighting scalar of the LQR. In order to demonstrate the improvement of the control efficiency, the optimally designed system was compared to a conventional design consisting of a cantilever aluminum beam with a PZT patch actuator at the root and 2 PVDF sensors fully covering the beam's underside. Simulation and experiment results are presented to show the effectiveness of the proposed design technique.
Collection of Technical Papers - 11th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
Optimal design methodology for an active structure using simultaneous optimization.
Collection of Technical Papers - 11th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference,
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/2452