Wavelet analysis was applied in two case studies. The first case is the signal analysis of a record of propeller-induced underwater pressure. The travelling bubbles collapsing noise was separated from the low frequency signal, which is related to the blade rate, by means of wavelet analysis and thereby could be studied in detail. The second case is about the separation of the acceleration due to the structural vibration from the rigid body acceleration recorded in full-scale trial of a planing boat in waves. The results from the two cases show that the wavelet analysis is an efficient tool for separating signals resulted from different mechanisms.
Fourier analysis is a widely applied tool for signal analysis, which transforms the time-domain signal into frequency-domain with assumption that the signal components are sinusoids of different frequencies. But not all signals satisfy this assumption. There are a lot of signals of non-sinusoid type. Sometimes it is interesting to study the detail signal characteristics It occurs sometimes that a measured signal consists of two or three kind sub-signals, which may be related to each other, or caused by different sources. The purpose of signal analysis is then to separate the sub-signals as completely as possible and study the details of each individual signal. In this context, Fourier analysis becomes less capable. Wavelet analysis is a windowing technique with variable-sized regions. The long time intervals can be used where the tow frequency information is desired, and shorter regions, where high frequency information is desired. One major advantage offered by wavelets analysis is the ability to perform local analysis- that is, to analyze a localized area of a larger signal.
