Fault Detection and Diagnosis of Rotary-Wing UAV's Rotor Using Multiclass Support Vector Machine

As the number of Unmanned Aerial Vehicles (UAVs) increases due to the development of modern technology, the safety of the UAVs should be guaranteed to accomplish their missions, In this study, a knowledge-based fault diagnosis method using Support Vector Machine (SVM) is proposed. The dimension of the state variables of the UAV is reduced utilizing the statistical feature of time-series data to configure the normal case and fault case data sets. Using SVM method, which is one of the supervised learning algorithms, the UAV is classified as a normal flight situation or a fault flight situation. In addition, multi classification is presented using both Root Mean Square (RMS) and Square Root of Amplitude (SRA) to distinguish which rotor is defective among four rotors. The numerical simulation is performed to verify the proposed fault diagnosis algorithm.