통합 데이터분석 알고리즘 기반 ROV 추진기 고장 진단 설계에 관한 연구

Abstract

Underwater thrusters are being used as driving elements for marine platforms such as underwater robots. Underwater thrusters can fault due to external or internal factors, causing economic losses or Accidents. In order to minimize these losses, a fault diagnosis system for underwater thrusters is essential. In this study, a Remotely Operated Vehicle (ROV) was designed, and an underwater thruster fault diagnosis system was studied based on the designed ROV. In this study, the broken thruster blade and entanglement of floating objects were selected as faults to be diagnosed. The two faults are difficult to distinguish through general analysis, and since the countermeasures for each fault are different, the algorithm was designed for the purpose of classifying the two faults. The ROV for this study implemented 6 degrees of freedom motion through 8 thrusters. The ROV shape takes symmetrical shape for stable thrust distribution and is divided into the main Cylinder and sub Cylinder. The main Cylinder consists of the ROV control board and sensors, and the sub Cylinder consists of a battery for power supply. The integrated data-based fault diagnosis algorithm used in this study uses AHRS gyro data to classify the underwater thruster state and, when a fault occurs, measures the change in ROV attitude. attitude changes predict the location of the fault thruster, and isolation and diagnosis of the fault thruster analyze current data and vibration data. The vibration data uses discrete Fourier transform to determine the type of fault through the analysis result of the rotation frequency and vibration magnitude in the frequency domain and the consumption current analysis. A performance test was conducted to verify the reliability of the sensor data for fault diagnosis, and the performance of the integrated data-based underwater thruster fault diagnosis algorithm was confirmed in the water tank.