In the paper, we discuss the use of the homogenized energy model (HEM) to develop a dynamic mode decomposition surrogate model for a PZT bimorph actuator used for micro-air vehicles including Robobee. The HEM quantifies the nonlinear, hysteretic, and rate-dependent behavior inherent to PZT in highly dynamic operating regimes. Due to the computation complexity of the HEM, we must develop a surrogate model. The surrogate model must be parameter- and control-dependent to be able to perform inverse problems or uncertainty quantification in different driving regimes. In the literature, DMD can be adapted to address different control inputs. We will discuss using interpolation over the parameters to adapt the DMD to include parameter dependence. Finally, we will discuss the results and limitations of the new surrogate model.
|