Structural instantaneous frequency extraction based on improved multi-synchrosqueezing generalized S-transform

Abstract

A new method is proposed to improve the accuracy of structural instantaneous frequency (IF) extraction. The proposed method combines a new form of improved generalized S-transform (IGST) and a multi-synchrosqueezing operation. The parameters selection of the window function in IGST is derived through the concentration measure (CM) principle. Then, the multi-synchrosqueezing algorithm is employed to improve energy aggregation of time-frequency analysis (TFA). To verify the effectiveness and accuracy of the proposed improved multi-synchrosqueezing generalized S-transform (IMSSGST), a frequency-modulated multi-component signal is investigated. For structural IF extraction, a two-story shear frame and a threestory steel frame structure are introduced. Furthermore, the IF identification of a seven-story RC shear wall structure is conducted to verified the practicability in actual engineering. Numerical simulation and experimental results show that the proposed method can effectively improve the energy aggregation of TFA and effectively improve the accuracy of IF identification.