14:15 〜 14:30
[S07-9-04] Processing Strategy On Strong Motion Records Of Bizarre Waveforms
Strong motion records are the first-hand data and precious indeed for research on earthquake engineering, which play an important role in improving structural seismic design, rapid earthquake-intensity reporting and earthquake early warning. However, there are some abnormal data in strong motion records, which reduces the application confidence of the observed data so as to affect the conclusion of scientific research. So, a rich set of strong motion records are analyzed in some typical earthquakes domestic and abroad in this paper. Four kinds of abnormal phenomena on the acceleration waveform are revealed, such as spike, asymmetric waveform, obvious baseline drift and strong motion records packets separation. Then reasonable recognition and processing approaches are derived from the preliminary analysis of the generation mechanism for abnormal phenomena. It is shown that spikes can bring ill effect on the rapid earthquake-intensity reporting and so on, so it should be identified with ‘jerk' method, ratio method and the consistency of the three-component PGA time, and then corrected with average value. For asymmetric waveform, the diversity mechanism is due to instrument pier, instruments and wrong operations. And the area average method and modified moving average method can be applied to identify acceleration records with the characteristics of asymmetric waveform. At the same time, the asymmetric waveform correction for strong motion records from Yibin Gaochang in Wenchuan earthquake is obtained by using the Butterworth low-pass filtering. In addition, two pieces of strong motion record packets can be connected by searching continuous and repeated data. Finally, the method of cumulative adding are used to find the clear baseline drift. At the same time, it is found that the abnormal waveform directly affects the characteristics of time history and frequency spectrum.