Analysis of vibration and sound signals captured by smartphone/tablet on board a helicopter
Helicopters are complex systems whose performance must be monitored continuously in order to guarantee the absolute imperative of flight safety. This leads to the need for a complete diagnosis of mechanical faults occurring in the gearboxes, blades, etc. These potential faults will be revealed essentially, through the analysis of acoustic signals recorded, on board or on the ground, by a cell phone or a portable recording tool, hence the originality of this thesis. These low-cost measurements, in addition to the complexity of the studied system, require processing methods capable of extracting the relevant diagnostic information. The vibration signals will complete this analysis by serving, for example, to estimate the instantaneous speed which is a priori, unknown. The proposed methodology should be robust and adapted to any operating regime. It should also be able to separate the different kinematic families existing in the signal. Then, it should be sensitive to the statistical nature of each mechanical fault by separating the deterministic phenomena from the random phenomena. Thus, the signal processing tools will be adapted to each type of fault in order to assess the health of the mechanical part linked to this signature extracted from the signal.