Correcting data from an unknown accelerometer using recursive least squares and wavelet de-noising

Non-linear finite element analyses of structures that are subject to seismic actions require high quality accelerogram data. Raw accelerogram data needs to be adjusted to remove the influence of the transfer function of the instrument itself. This process is known as correction. Unfortunately, information about the recording instrument is often unknown or unreliable. This is most often the case for older analogue recordings. This paper uses a recursive least squares (RLS) algorithm to identify the instrument characteristics even when completely unknown. The results presented in the paper implement a modern approach to de-noising the accelerogram by employing the wavelet transform. This technique removes only those components of the signal whose amplitudes are below a certain threshold and is not therefore frequency selective. It supersedes to some extent conventional band pass filtering which requires a careful selection of cut-off frequencies, now unnecessary.