Reduced-order variational mode decomposition to reveal transient and non-stationary dynamics in fluid flows

A novel data-driven modal analysis method, reduced-order variational mode decomposition (RVMD), is proposed, inspired by the Hilbert–Huang transform and (VMD), to resolve transient or statistically non-stationary flow dynamics. First, form of RVMD modes (referred as an ‘elementary low-order dynamic process’, ELD) constructed combining representation idea intrinsic function, which enables computed characterize properties space–time fluid flows. Then, algorithm designed based on VMD achieve a low-redundant adaptive extraction ELDs in data, with solving elaborate optimization problem. Further, combination Hilbert spectral leads modal-based time-frequency framework view, providing potentially powerful tool discover, quantify analyse dynamics complex problems. To provide comprehensive evaluation, computational cost parameter dependence are discussed, well relations between some classic methods. Finally, virtues utility demonstrated via two canonical problems: cylinder wake planar supersonic screeching jet.