Modeling hydroelectric actuators with memory on spectral submanifolds

This abstract has open access
Summary
We present experiments on a new hydraulically amplified self-healing electrostatic (HASEL) actuator and propose a method for modeling its forced response on data-driven spectral submanifolds (SSMs). First, we find that the fast periodic dynamics is captured by a two-dimensional forced SSM. Second, over longer timescales and slower excitation, we discover hysteretic behavior and creep, and propose a new method for modeling this behavior on hyperbolic invariant manifolds. Our method, based on geometric singular perturbation theory, demonstrates for the first time how to account for memory effects in SSM-based modeling and provides a low-dimensional model suitable for control.
Abstract ID :
98
12 visits