A Space-Frequency Data Compression Method for Spatially Dense Laser Doppler Vibrometer Measurements

Abstract

When spatially dense mobility shapes are measured with scanning laser Doppler vibrometers, it is often impractical to lise phase-separation modal parameter estima­ tion methods due to the excessive number of highly coupled modes and to the prohibi­ tive computational cost of processing huge amounts of data. To deal with this problem, a data compression method llsing Chebychev polynomial approximation in the fre­ quency domain and two-dimensional discrete Fourier series approximation in the spatial domain, is proposed in this article. The proposed space-frequency regressive approach was implemented and verified using a numerical simulation of a free-free­ free-free sllspended rectangular aluminum plate. To make the simulation more realis­ tic, the mobility shapes were synthesized by modal superposition using mode shapes obtained experimentally with a scanning laser Doppler vibrometer. A reduced and smoothed model, which takes advantage of the sinusoidal spatial pattern of structural mobility shapes and the polynomial frequency-domain pattern of the mobility shapes, is obtained. From the reduced model, smoothed curves with any desired frequency and spatial resolution can he produced whenever necessary. The procedure can he used either to generate nonmodal models or to compress the measured data prior to modal parameter extraction. © /996 John Wiley & Sons, inc.