SYstem for Monitoring and Prevention for Helicopters Operating Non-stationary Conditions with Embedded Intelligence – SYMPHONIE

The aerospace industry has already made significant efforts to enhance energy efficiency and reduce its environmental footprint, through optimizations of dynamic transmission systems and hybridization using electric machines, which are also crucial to monitor, with a strong objective of reducing operational costs through a condition-based maintenance strategy.
To advance in this area, AIRBUS Helicopters aims to extend maintenance intervals or replace them with automated monitoring, the robustness of which must meet the requirements of certification authorities (EASA, FAA, etc.) and be insensitive to operational conditions (flight dynamics and maintenance actions).
To this end, AIRBUS Helicopters is considering other types of measurements, such as instantaneous speed, which is underutilized in aerospace but has shown relevant results in the field of conveyor wind energy. Monitoring electric machines is seen as an opportunity to explore new monitoring methods, particularly through currents and voltages, thus contributing to their diagnosis as well as that of the complete transmission system.
All these measurements—vibrations, instantaneous speeds, and electrical measurements—are strongly correlated through the power of the studied machines. It seems pertinent to develop a joint study program of these measurements to improve existing detection tools and evolve them into diagnostic tools for the helicopter's dynamic power transmission system.
Therefore, AIRBUS Helicopters has proposed this program to LASPI, with which it is already collaborating on the vibrational monitoring of complex systems. Given the ambition of this project, LASPI has called upon its partners LVA and LAMCOS from INSA Lyon, who possess highly complementary expertise in the field.
This program plans to enhance existing monitoring algorithms, particularly with an automatic adaptation of their parameters to generalize them across the helicopter fleet and make them less sensitive to operational conditions and maintenance actions. This work involves the development of advanced signal processing tools as well as the use of artificial intelligence for more efficient optimization of indicator parameterization. It also requires dynamic modeling that would explain certain phenomena, such as the disappearance of signals that are not yet understood, particularly in epicyclic reducers. This modeling work should allow for a better understanding of the modifications of transfer functions based on usage conditions and consequently the impact on various vibrational, speed, and current measurements.
To successfully carry out this project, experiments on dynamic test benches at AIRBUS Helicopters are necessary, as well as at LASPI on benches with epicyclic reducers and axial flux electric machines.
The results of this work will lead to the development of a library of software bricks and demonstrators to highlight their performance. The methodologies may be subject to publications, and the adaptation of these bricks to the industrial context will lead to tools at TRL6, aiming for a transition from research activities to industrial development for short-term integration into the HUMS system (onboard computer and ground station).
Furthermore, this research area is part of the training program for a master's course offered in Roanne, and the test benches and disclosable results will be integrated into this training program in the short term.