Predicting multi-parametric dynamics of an externally forced oscillator using reservoir computing and minimal data

• DOI: 10.1007/s11071-024-10720-w
• Publication Date: 2024-12-05T07:39:08Z
• Authors: Manish Yadav; Swati Chauhan; Manish Dev Shrimali; Merten Stender
• Keywords: Reservoir computing; Dynamics; Parametric oscillator

Abstract Mechanical systems exhibit complex dynamical behavior from harmonic oscillations to chaotic motion. The dynamics undergo qualitative changes due internal system parameters like stiffness and external forcing. Mapping out complete bifurcation diagrams numerically or experimentally is resource-consuming, even infeasible. This study uses a data-driven approach investigate how bifurcations can be learned few response measurements. Particularly, the concept of reservoir computing (RC) employed. As proof concept, minimal training dataset under resource constraint problem Duffing oscillator with forcing provided as data. Our results indicate that RC not only learns represent for seen during training, but it also provides qualitatively accurate robust predictions completely unknown multi-parameter regimes outside while being trained solely on regular period-2 cycle dynamics, proposed framework correctly predicts higher-order periodic out-of-distribution signals.