Closely Spaced Mode Identification of Skyscraper Buildings via Variational Mode Decomposition with Multiple-Signal Classification Algorithm and Recursive Hilbert Transform

Addressing the limitations of traditional modal analysis methods, which struggle with closely spaced modal frequencies and noise interference, this paper introduces a novel approach that integrates variational mode decomposition (VMD) with multiple signal classification (MUSIC) and the recursive Hilbert transform (RHT). This integration leverages the adaptability of VMD in signal decomposition and exploits the high-resolution spectral identification capabilities of MUSIC to decompose the closely spaced modes accurately. Additionally, the RHT is applied to analyze the instantaneous properties of the decomposed signals. The proposed method was validated through a numerical study on a 2DOF model, demonstrating its effectiveness and robustness in identifying modal parameters under simulated conditions. Further validation was conducted through field measurements from a 420[Formula: see text]m high skyscraper building during Super Typhoon Nesat, confirming the practical applicability and effectiveness of the approach in actual engineering cases. The findings indicate a substantial improvement in the accuracy of closely spaced modal parameter identification, thus enhancing the reliability of structural health monitoring (SHM) systems in skyscraper buildings.