VIBRATION SIGNAL ANALYSIS AS SHIP'S DIAGNOSTIC'S TOOL

Condition monitoring is an invaluable measure of modern engine diagnostic tool techniques. Vibration signal is a basic signal source, which is analyzed by signal analysis techniques. This paper presents an overview of wavelet application in ship’s engine condition monitoring. Due to a nature of the vibration signal (non-stationary in origin), wavelet and related transforms seems ideally suited to vibration signal analysis. The paper deals with cutting-edge technologies in this research field, as integrated vehicle health monitoring, Bayesian networks approach (which became superior to both the fault tree analysis and the event tree analysis), wavelet learning techniques, etc. Examples of typical failures are shown, as well as the effect of the failures to the vibration signal. These effects make possible to develop and use an identification procedure, which leads to the localization of the failure.