Damage identification of multirotor UAV propellers via unsteady coupling association

To ensure the operational safety of unmanned aerial vehicles (UAVs) and to reduce maintenance costs, real-time monitoring of their system statuses is essential. However, research on multirotor propeller fault diagnosis is limited, and no standardized methodology exists for classifying and evaluating propeller faults. This study proposes a propeller fault diagnosis method based on pressure coupling of pressure pulsation and vibration signals on the arms of multirotors. Aerodynamic models of three types of damaged propellers were established, and the pressure pulsation phenomenon caused by rotating propellers on the arms was investigated using computational fluid dynamics (CFD) methods. Subsequently, a propeller damage identification sensing system was developed, and a propeller damage identification model based on a Long Short-Term Memory (LSTM) network was constructed. The model achieved an accuracy of 96.75%, providing a new approach for improving the accuracy of propeller fault monitoring in multirotors. • Developed aerodynamic models for propellers with three types of damage. • CFD observed lift and pressure changes from propeller damage, guiding sensor setup. • Constructed a coupled association propeller damage identification sensing system. • A propeller damage identification model based on LSTM was proposed.