Prognostics of aerospace electromechanical actuators: comparison between model-based metaheuristic methods

Abstract

Abstract Electro-Mechanical Actuators (EMAs) deployment as aircraft flight control actuators is an imperative step towards more electric concepts, which propose an increased electrification in aircraft subsystems at the expense of the hydraulic system. Despite the strong benefits linked to EMAs adoption, their deployment is slowed down due to the lack of statistical data and analyses concerning their often-critical failure modes. Prognostics and Health Management (PHM) techniques can support their adoption in safety critical domains. A very promising approach involves the development of model-driven prognostics methodologies based on metaheuristic bio-inspired algorithms. Evolutionary (Differential Evolution (DE)) and swarm intelligence (particle swarm (PSO), grey wolf (GWO)) methods are approached for PMSM based EMAs. Furthermore, two models were developed: a reference, high fidelity model and a monitoring, low fidelity counterpart. Several failure modes have implemented: dry friction, backlash, short circuit, eccentricity and proportional gain. The results show that these algorithms could be employed in pre-flight checks or during the flight at specific time intervals. Therefore, EMA actual state can be assessed and PHM strategies can provide technicians with the right information to monitor the system and to plan and act accordingly (e.g. estimating components Remaining Useful Life (RUL)), thus enhancing the system availability, reliability and safety.