Page 19 - Vol.10
P. 19

Figure 10. Results after baseline correction


                   0.5
                                                     Acceleration: after baseline correction
                   0.3
                   0.1
                  g
                   -0.1
                   -0.3

                   -0.5
                      0            10            20            30            40            50            60
                                                              second
                   0.2
                                                      Velocity: after baseline correction
                   0.1

                  m/s  0

                   -0.1

                   -0.2
                      0            10            20            30            40            50            60
                                                              second
                  0.003
                                                    Displacement: after baseline correction
                  0.002
                  0.001

                 m  0
                 -0.001

                 -0.002
                 -0.003
                      0            10            20            30            40            50            60
                                                              second







              –  Least square straight line fit for   domain convolution operates, first   The time-domain convolution of
                d 0 (t): d 0 +f 0 (t)          we copy and inverse the original   Ormsby filter with Kaiser window
              –  C o r r ec t ed  d is p l a c e m e n t   time history to negative time to have   and baseline corrected acceleration
                d 1 (t)=d 0 (t)-d 0 -f 0 (t)   a symmetric time history, then we   is shown in Figure 11, the waveform
              –  Low – pass Ormsby filter for d 1 (t)   start time domain convolution using   show there’s low – frequency content
                to get d 2 (t)                 the MATLAB algorithm, but the   by low–pass Ormsby filter, it’s
              –  Final  corrected  displacement   final results will be truncated from   easier to understand in Figure 12
                d 3 (t)=d 1 (t)-d 2 (t)        the point that the peak of Ormsby   The final results illustrated in Figure
                                               filter at the time zero.
              No w  s e t  t he  O r m sby  c or ne r                          13 show the amplitude is smaller
              frequency fc as 2.8 Hz and ft as   Figure 10 shows the results after   than the original one because lot of
              3.2 Hz, and Kaiser window with   baseline correction, the velocity is   low frequency content is ‘filtered’
              α=10. Figure 8 illustrates the Ormsby   integrated from the acceleration   which is an important portion of
              filter, Kaiser window and the    wh ich has been cor rected by   earthquake wave, which can be
              multiplication of Ormsby filter and   base l i ne cor rect ion; a lso  the   observed in Figure 12 that the portion
              Kaiser window, the time window is   displacement is integrated from   below 3 Hz (the Ormsby frequency
              two seconds.                     corrected velocity. The results are   given in this example) plays an
              Figure 9 descripts how the time-  more reasonable.               important role for this earthquake.




                                                                              NEW FAB TECHNOLOGY JOURNAL         JUNE  2013  19
   14   15   16   17   18   19   20   21   22   23   24