Fuzzy-Oriented Anomaly Inspection in Unmanned Aerial Vehicle (UAV) Based on MEMS Accelerometers in Multimode Environment

Abstract

Apart from brushless dc (BLdc) motors, structural components that makeup the unmanned aerial vehicle (UAV) are also critical. Failure in the UAV’s structure may put the UAV at risk and eventually lead the UAV to crash. In this study, a real-time anomaly inspection of the UAV, focusing on the UAV’s arms structure, is proposed. A crack in the UAV’s arm is investigated at different UAV modes, and a fuzzy logic algorithm will determine the UAV’s condition based on the vibration level observed. Experimental studies show that the UAV’s arms with minor and major cracks generate approximately 38%–50% and 62%–71% maximum amplitude compared to a healthy arm in ground mode. The UAV can still take off and hover with a minor cracked arm, whereas a major crack in the UAV’s arm will make the UAV crash. In the flight mode, the vibration amplitudes exerted by the UAV with minor cracked arms are in the range of 1.53–1.74 g, approximately a 15%–32% increase compared with the healthy arm. The proposed method is practical to be applied on top of the UAV as the overall weight is only 265.3 g, and the UAV’s condition can be monitored through a smartphone.