Model-Free Sampling Method for Basins of Attraction Using Hybird Active Learning
The proposed sampling method is comprised of: 1) additional sampling on trajectories, 2) margin-based active learning, and 3) density-based sampling.
Attractor Selection of Nonlinear Energy Harvesting Using Reinforcement Learning
Several control policies learned by a reinforcement learning method (DDPG) realize nonlinear energy harvesters’ operating on desired attractors.
Attractor Switching for Nonlinear Dynamical Systems Using Deep Reinforcement Learning
Two reinforcement learning algorithms (DDPG and CEM) were implemented to realize switching between coexisting attractors in nonlinear dynamical systems.
Dynamics of a Non-Contact Translational-to-Rotational Magnetic Transmission
This system exhibits nonlinear behaviors, such as harmonic, sub-harmonic, and chaotic responses, when the external excitation is sinusoidal.
Dynamics of Unforced and Vertically Forced Rocking Elliptical and Semi-Elliptical Disks
The rocking disk can be mono-stable or bi-stable by altering its geometry. Its dynamic response exhibits multiple periodicities under vertical excitations.