In the scope of the development of a more electrical aircraft, the project aimed to develop a reliable tool for the health monitoring of electro-mechanical actuators (EMAs). The replacement of hydraulic systems with the implementation of EMAs on flight control actuation systems of both A/C and H/C can lead to many advantages but, since the mechanical jam of such a system has to be considered as a possible and high critical event, research activities have to pay attention to the compliance with the operational reliability and safety requirements.
One of the most attractive and promising way to meet these requirements is the implementation of advanced health monitoring systems. Use of sensors for the detection of failures has been successfully proved by the research community, including Umbra, but the challenge of the HOLMES project is to rely on components that are basic on an EMA like electric motors, position sensors, motor commutation devices without additional sensing on the internal mechanical components (vibration, temperature, etc). A promising parameter is the phase electric motor currents but other (e.g. voltage, speed of the motor and position of the actuator) will be evaluated. When this is the case, the reliability and safety of the EMA will be improved without any penalty in terms of weight and envelope.
So, the strategy of the HOLMES project was to simplify the hardware measurement layout, moving complexity in the software and algorithm design. In this view, a mixed approach for FDI was proposed. When a degradation (in a stochastic sense) of “on flight” measured variables were evident with respect to the nominal values, a model-free FDI generates an alarm. This information was integrated by the results of a spectral analysis of the measured signals that put in evidence the presence of unexpected harmonics. The two sources of information were merged by using black box identification methods.
The problem of detecting faults in electromechanical actuators (EMA) has been tackled by many scientists over the last 20 years and some researches can be found in literature.
The replacement of hydraulic systems with the implementation of EMAs on flight control actuation systems of both A/C and H/C can lead to many advantages but, since the mechanical jam of such a system has to be considered as a possible and high critical event, research activities have to pay attention to the compliance with the operational reliability and safety requirements.
Two major approaches have been pursued: model based and model free methods. Model-free methods are preferred than model-based methods, since the latter required a mathematical model of the system than it is often difficult to design with sufficient accuracy. The model-free methods are based only on signal processing of measured variables and they are focused on the goal of rising alarms and classifying the faults. However, the trend in current research is to combine the two methodologies, to gain the benefits of both worlds.
Such later approach was demonstrated to be promising for the early identification of mechanical failure of ballscrews, gears and bearings. Such results have to further developed in order to be used in actual applications.