This work describes a novel method for experimentally measuring and identifying dynamic parameters of bistable systems. The investigated samples are physical models of unit cells of mechanical metamaterials used for seismic energy dissipation. In order to identify this type of nonlinear systems, the paper proposes a new instantaneous probabilistic algorithm, which is capable of extracting the causal components of the dynamic response from the random ones. This is especially relevant for nonlinear systems exhibiting abrupt transitions between dynamical regimes, whose frequency content can evolve significantly over time. In particular, the approximate Bayesian computation (ABC) approach is extended to the time-frequency representation of non-stationary dynamic responses typical of bistable systems. The method was tested on experimental measurements of 3D-printed bistable samples with the final goal of estimating the instantaneous energy dissipated under different external excitations, proving the device's effectiveness.
Scussolini et al. (Tue,) studied this question.