Answers are grounded in this paper and cite their sources.
This paper proposes a two-dimensional asymmetrically coupled multistable stochastic resonance system driven by cross-correlated Ornstein–Uhlenbeck colored noise to address blurred time-frequency ridges, weak feature extraction difficulty, and insufficient recovery accuracy in nonstationary signals under strong background noise. The proposed system enhances weak nonstationary signals through an asymmetric potential landscape, a two-dimensional coupling structure, and multistable transition channels. First, a two-dimensional asymmetrically coupled multistable potential function is constructed, and its critical points, potential well–saddle topology, and multichannel transition structure are analyzed. Then, an analytical framework involving the effective potential function, stationary probability density, probability flow, mean first-passage time, and multistate master equation is established based on the unified colored-noise approximation. This framework reveals the effects of cross-correlated colored noise, barrier height, and channel transition rate on stochastic resonance enhancement. Finally, multicomponent nonstationary signals and variable-speed nonstationary signals are used for validation, and the system performance is evaluated under different noise intensities, different model comparisons, ablation experiments, and time-domain recovery errors. The results show that the proposed method effectively enhances target time-frequency ridges and suppresses background interference. Compared with the one-dimensional multistable system, the symmetric two-dimensional coupled system, and the non-stochastic-resonance EMD method, the proposed method achieves better ridge continuity, energy concentration, and waveform recovery accuracy. This study provides a new nonlinear dynamic approach for weak nonstationary feature extraction and signal recovery in complex mechanical systems under strong noise backgrounds.
No takes yet. Share an insight, caveat, or question.
Yang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: