Diagnostics of electro-mechanical actuators based upon the back-EMF reconstruction

Electrical systems are gradually replacing the more traditional hydraulic and pneumatic solutions for the transmission of secondary energy for onboard aircraft equipment. Therefore fault detection and health management strategies properly conceived for electrical devices are becoming a highly relevant topic for research and development in the aerospace disciplines. One possible practical implementation of these methodologies would be the identification of parameters for diagnostic and prognostic monitoring, which are highly sensitive to incipient faults but, at the same time, are less influenced by operating conditions (external loads, command input, temperatures, etc.). In this paper, the authors evaluated the effectiveness of counter-electromotive force (back-EMF) coefficient as a prognostic parameter, emphasizing a novel sampling approach that significantly lower the computational effort required while maintaining a good back-EMF coefficient curve reconstruction. The approach is virtual sensor-like, using only already available data for the correct operation of the BLDC motor. The proposed method was tested by evaluating the back-EMF coefficient reconstruction as a function of some progressive failures typical of EMA motors, such as inter-turn partial shorts and rotor static eccentricity. Its robustness to external disturbances has been tested by evaluating different actuation commands and operating conditions. As expected, the back-EMF signal shows a marked dependence on the considered failure modes and, at the same time, a suitable insensitivity to the other external factors.