Predictive Maintenance Optimization for Aircraft Redundant Systems Subjected to Multiple Wear Profiles

Researchers and industry practitioners have recently demonstrated great interest in maintenance planning since it directly impacts the lifecycle cost and availability of systems. Considering the aviation industry, this impact is even more relevant due to high-availability expectations from aircraft operators and high costs incurred when an aircraft becomes out of service. For this reason, some minor maintenance activities are carried out near the gate, between two consecutive flight legs. These activities are referred to as aircraft line maintenance. The usage of prognostics and health management techniques may help the prevention of unexpected events when caused by degrading components. The application of prognostics techniques together with a line maintenance planning optimization approach is not a trivial task and requires sophisticated mathematical approaches especially when dealing with complex and integrated systems. This paper proposes a methodology for predictive line maintenance optimization of redundant aeronautical systems subjected to multiple wear conditions. Degradation trends and future wear values are estimated considering an implementation of a multiple model approach of the extended Kalman filter technique. Planning optimization is based on the minimization of operational costs comprising several aspects of the aviation industry such as dispatch requirements, delays, cancellations, and equipment costs. A case study is conducted using field prognostics data of hydraulic systems to evidence the efficiency and benefit of the proposed methodology.