Detection of mechanical fault in linear servo motion systems using synchronous-averaging and wavelet de-noising

Machinery condition monitoring by vibration analysis has been discussed over decades. Varieties of rotor, bearing and gearbox diagnostic techniques are available. Most of them are developed for rotary machines. Traditional spectrum analysis can successfully detect defects induced from a periodical impact signal under constant speed operation. However, a linear motion system does not exhibit a periodic behavior and the diagnosis technique is relatively new for researchers. This paper introduces a synchronous averaging and wavelet de-noising combined technique to detect defects on a precision linear motion servo-control system. Experimental results of the proposed method are also presented.