**5. Conclusions**

This chapter deals with WT applications for practical condition monitoring issues on flexible couplings and i.c. engine. In particular, CWT and DWT capability was assessed. The former was used for faulty event identification and impulse event characterization through the analysis of three-dimensional representations of CWT coefficients. The latter was applied for filtering and feature extraction purposes and for detecting impulsive events which were strongly masked by noise. Several CWT representation improvements were also evaluated.

Comparing the results from both the CWT and DWT analyses, the ability of WTs in satisfying both condition monitoring and fault detection requirements for all tested cases was clearly demonstrated. In particular, traditional CWTs of the residual signal (i.e. the signal obtained by subtracting the time synchronous average from the raw signal) with the Morlet mother wavelet was revealed to be the most powerful tool in angularly localizing the assembly fault within the engine kinematics.

It can be concluded that the application of WTs not only enables changes in the state of the tested machine to be recognized but also the localisation of the source of the alteration.
