Multivariate empirical mode decomposition (MEMD) method is explored in this paper to perform modal identification of structures using the multisensor vibration data. Due to inherent sifting operation of empirical mode decomposition (EMD), the traditional MEMD results in mode-mixing that causes significant inaccuracy in modal identification and condition assessment of structures. Independent component analysis, another powerful blind signal decomposition method, is integrated with the MEMD to alleviate mode-mixing in the resulting modal responses. The proposed technique is verified using a suite of numerical, experimental, and full-scale studies (a building tower in China and a long-span bridge in Canada) considering several practical applications such as low-energy frequencies, closely spaced modes, and measurement noise. The results confirm the improved performance of the proposed method and prove that it can be considered as a robust system identification tool for flexible structures.Out of several time-frequency analysis methods, WT is a linear transformation of a signal using an appropriate basis function that is localized in both frequency and time domain. [4,8] It results in accurate representation of vibration measurement because of its compactness, quick decaying characteristics, and higher order moments. Unlike WT, BSS [21] decomposes original source signals from their observed mixed signals without knowing system information and the source signals. The BSS method is free of basis functions and used in many different fields of science and engineering. In general, there are two popular tools to solve the BSS problems-independent component analysis (ICA) and second-order blind identification. Although the ICA method utilizes high-order statistics to separate the inherent signals, the second-order blind identification method uses the second-order statistics like covariance matrices of vibration measurements for the decomposition. [13] EMD is one of time-frequency decomposition tools that can deal with both nonlinear and nonstationary signals [22] without requiring any basis functions. Owing to this property, it has gained significant popularity in the area of structural condition assessment and damage detection. The EMD method was originally developed by Huang et al. [20] that creates a set of intrinsic mode function (IMF) representing time and frequency domain information of the signal. The EMD is also known as an adaptive time-frequency decomposition method that decomposes a multicomponent signal into a number of mono-component signals. The IMFs are extracted by sifting operation that is undertaken by multiple averaging and interpolation operations of the raw signal. However, sifting operations cause considerable mode-mixing in the IMFs. [23] Recently, a wavelet-bounded EMD is proposed to find an optimization-based solution to the problem of mode-mixing. [24] Unlike the WT and BSS, the EMD method is self-adaptive in nature. [20] Furthermore, the EMD method can deal with a single real-valued signal to obta...
