Extraction of the Bridge Temperature Strain Effect Based on EMD and IMF Energy

To extract the temperature effect component of the dynamic strain signal through empirical mode decomposition (EMD), which presents the shape of a cycle trend measured from a bridge, the intrinsic mode function (IMF) order number by which temperature strain is constructed should be ascertained. The problem is solved on the basis of the selection of the IMF order threshold. The energy of signal of the IMF is analyzed in the time and frequency domains, and the threshold of the IMF order is acquired through a synthetic analysis of the energy catastrophe order and the correlation coefficient of the Hilbert marginal spectrum. Finally, the temperature effect component of the strain data is extracted by selecting the cycle trend of the IMF according to the IMF order threshold. The result shows that (1) the periodic trend component of a dynamic strain leads to the order catastrophe of the IMF energy from which the value range of the IMF order threshold is preliminarily determined; (2) the correlation coefficient of the IMF marginal spectrum increases rapidly to over 0.8 by analyzing the energy distribution in the frequency domain of the IMF located in a preliminarily determined range. The energy distribution patterns of last-order IMFs are consistent. By applying the method to the in-situ data, the result demonstrates that the IMF order threshold can be precisely obtained by the first correlation coefficient of the IMFs marginal spectrum of over 0.8. After the temperature effect component is extracted based on the threshold, the live load information in the dynamic strain is properly retained.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01600468
  • Record Type: Publication
  • Files: TRIS, ASCE
  • Created Date: May 10 2016 3:02PM