Electromechanical Actuators Affected by Multiple Failures: Prognostic Method based on Wavelet Analysis Techniques

Incipient failures of electromechanical actuators (EMA) of primary flight command, especially if related to progressive evolutions, can be identified with the employment of several different approaches. A strong interest is expected by the development of prognostic algorithms capable of identifying precursors of EMA progressive failures: indeed, if the degradation pattern is properly identified, it is possible to trig an early alert, leading to proper maintenance and servomechanism replacement. Given that these algorithms are strictly technology-oriented, they may show great effectiveness for some specific applications, while they could fail for other applications and technologies: therefore, it is necessary to conceive the prognostic method as a function of the considered failures. This work proposes a new prognostic strategy, based on artificial neural networks, able to perform the fault detection and identification of two EMA motor faults (i.e. coil short circuit and rotor static eccentricity). In order to identify a suitable data set able to guarantee an affordable ANN classification, the said failures precursors are properly pre-processed by means of Discrete Wavelet Transform extracting several features: in fact, these wavelets result very effective to detect fault condition, both in time domain and frequency domain, by means of the change in amplitude and shape of its coefficients. A simulation test bench has been developed for the purpose, demonstrating that the method is robust and is able to early identify incoming failures, reducing the possibility of false alarms or non-predicted problems.