Influence of parameters selection in Chebyshev filters on the strong motion data processing

Abstract

This contribution discusses the application of Chebyshev Type I filter for processing real earthquake records. Consideration is given to the effects of filtering parameters (passband amplitude ripple and order of the filter) on the time series, strong-motion parameters, Fourier Amplitude Spectrum of acceleration, and elastic displacement response spectra. Time histories of five earthquakes with different moment magnitudes have been examined (from stations located close to the epicenters). Data processing is based on application of bandpass Chebyshev filtering over frequency range with substantial signal to noise ratio (level of 3 or approximately 3 dB). Applying different filters, we have monitored several important strong-motion parameters: peak values of acceleration, velocity, and displacement; Arias intensity, acceleration/velocity spectrum intensity, significant duration, etc. Some new results and conclusions concerning the influence of Chebyshev filter in data processing of records have been summarized. The graphical and numerical outcomes obtained, as well as the comparison with a Butterworth causal filter, are included in the work. The results could be potentially useful to engineering seismologists who need to evaluate and better understand the merits of this type of filtering for strong-motion data processing.

Three-dimensional dynamic deflections and natural frequencies of a stiff footbridge based on measurements of collocated sensors

“…At a second step, to remove the remnant noise and identify the oscillation amplitude, GNSS data were filtered using two different filters, the k‐filter, a filter derived from supervised learning experiments , and the Chebyshev filter , and the output of the two methods was compared.…”

“…Because GNSS short-period component presents a large amount of noise, permitting to identify oscillation frequencies but not the peak oscillation amplitude, GNSS measurements were further filtered as described in the following paragraphs. At a second step, to remove the remnant noise and identify the oscillation amplitude, GNSS data were filtered using two different filters, the k-filter, a filter derived from supervised learning experiments [28], and the Chebyshev filter [16,20,47,48], and the output of the two methods was compared.…”

Monitoring horizontal displacements in a vertical profile of a tall industrial chimney using Global Positioning System technology for detecting dynamic characteristics

“…The main reason for filtering of the GPS recordings was to de‐noise displacement records. The efficiency of the used filter processing was also confirmed in the papers .…”

Μέτρηση υψίσυχνων ταλαντώσεων με GPS δειγματοληψίας 100Hz

“…Ένα μειονέκτημα των φίλτρων Chebyshev είναι η κυμάτωση στη ζώνη διέλευσης, το οποίο εξασφαλίζει όμως τη μικρότερη δυνατή ζώνη μετάβασης (transition band) (Hamming, 1998). Στην παρούσα διατριβή ο σχεδιασμός του φίλτρου έγινε με έναν σχετικά μικρό συντελεστή εξασθένισης (0.5dB) ώστε να συνδυάζεται κατά βέλτιστο τρόπο το μικρό εύρος ζώνης απόκρισης και μικρή κυμάτωση στη ζώνη διέλευσης (Smith, 1997;Mollova and Scherbaum, 2007).…”

“…Ειδικότερα για τη δεύτερη περίπτωση σε πολλές μελέτες ταλαντώσεων κατασκευών αποκαλύπτεται η ύπαρξη ιδιοσυχνοτήτων με παραπλήσια τιμή η οποίες είτε αντιστοιχούν σε διαφορετικές ιδιομορφές (Papageorgiou and Lin, 1989;Moschas and Stiros, 2014) είτε είναι αποτέλεσμα μεταβολής της κύριας ιδιοσυχνότητας λόγω της φόρτισης, λόγω πλαστικοποίησης της κατασκευής κτλ (Durukal et al, 2003;Macdonald and Daniell, 2005;Soyoz and Feng, 2008 αντιστοιχούν σε διαφορετικές ιδιομορφές (Papageorgiou and Lin, 1989;Moschas and Stiros, 2014) είτε είναι αποτέλεσμα μεταβολής της κύριας ιδιοσυχνότητας λόγω της φόρτισης, λόγω πλαστικοποίησης της κατασκευής κτλ (Durukal et al, 2003;Macdonald and Daniell, 2005;Soyoz and Feng, 2008 αντιστοιχούν σε διαφορετικές ιδιομορφές (Papageorgiou and Lin, 1989;Moschas and Stiros, 2014) είτε είναι αποτέλεσμα μεταβολής της κύριας ιδιοσυχνότητας λόγω της φόρτισης, λόγω πλαστικοποίησης της κατασκευής κτλ (Durukal et al, 2003;Macdonald and Daniell, 2005;Soyoz and Feng, 2008 (Boore and Bommer, 2005;Boore, 2005;Emore et al, 2007;Mollova and Scherbaum, 2007). (Dallard et al, 2001;Roberts et al, 2006 (Wacker et al, 1998;Custódio et al, 2009;Τριανταφύλλου, 2013).…”