Reconstruction of compressed signals for bridge structures based on autoregressive iterations
LI Cui
HUANG Kan
LIAO Yanfu
ZHU Daopei
Abstract：[Objective]The vibration signals of bridge structures under ambient excitation usually present low signal-to-noise ratio and poor sparsity,which makes it difficult to realize high-precision compressed sensing.To solve this problem,a compressed signal reconstruction method based on autoregressive iteration was proposed.[Methods]The unknown reconstruction error was characterized by the autoregressive error,the autoregressive coefficients were taken as the optimization variables,and the minimization of autoregressive error was set as the optimization objective.An autoregressive iterative calculation model for signal reconstruction was established,and the measured signals of an actual bridge were adopted for calculation and analysis.[Results]The results show that the autoregressive iteration has the characteristic of transferring high-frequency energy to the low-frequency band.The minimum reconstruction error can be obtained only when the autoregressive errors of the reconstructed signal and the original signal are close.Compared with the orthogonal matching pursuit method,the proposed method achieves significantly higher reconstruction accuracy of compressed sensing.For the original signal,the minimum reconstruction error appears at the second iteration step under different model orders;for the filtered signal,the minimum reconstruction error can be acquired after iterative convergence with a reasonable combination of compression-sampling ratio and cutoff frequency.The proposed method can be well applied to the compressed sensing reconstruction of vibration signals with low sparsity.
Keywords：Structural health monitoringCompressive sensingBridge structureSignal reconstructionAutoregressive model
Publication Date：2026-05-31
Online Publishing Date：2026-09-12(First online date of this platform, not the publication date of the document)
Pages：9( 150-158 )