New Lumped Parameter Numerical Model For Monitoring Electromechanical Actuators

The increasing use of all-electric secondary power sources in onboard systems is leading to the increased use of ElectroMechanical Actuators (EMAs) in the aerospace industry. However, ensuring acceptable levels of safety and reliability requires the development of new prognostic and diagnostic methods to detect faults early and prevent the degradation of EMA performance. Adopting a model-based approach proves beneficial to designing various algorithms tailored to specific purposes. These models range from simplified monitoring models that balance computational efficiency with precision, to high-fidelity models that accurately simulate system behavior. High-fidelity models are used to generate databases, create predictive algorithms, and train machine-learning surrogates. This study introduces a Simulink Medium Fidelity (MF) model for a 3-phase PMSM-driven (Permanent Magnets Synchronous Motor) aeronautical servoactuator. Positioned between the High Fidelity (HF) and Low Fidelity (LF) models in terms of structure, detail, and accuracy, the new model enables simulation of the operation of a three-phase PMSM motor without relying on the simplifications found in typical LF models. For example, it avoids reduction to an equivalent single-phase motor model or linearization of the stator circuit. Despite its lower computational cost, the Medium Fidelity (MF) model maintains consistency with High Fidelity (HF) models. It considers the mutual interaction between different phases and potential imbalances caused by saturation in an inverter Hbridge branch when solving the electromagnetic circuit of the stator. Despite these considerations, the MF model demonstrates commendable computational performance and accuracy.