초음파 신호의 특성을 이용한 기어박스의 결함 분류

Abstract

In the development of a fault diagnosis and condition monitoring in the gearbox, the research is a quantitative analysis and a test of gear damage effect on the vibration of the gearbox. The Lab-scale gearbox test device that builds the several types of fault such as gear tooth breakage, misalignment and bearing fault occurred by gearbox fault simulator. This paper presents feature analysis through the GA (genetic algorithm) and SVM (support vector machine) of machine learning, the performance of feature classification is evaluated by faults on the gearbox. The research compared the possibility of classification through of the vibration signal and the ultrasonic signal by using the features applicable to the gear box was confirmed and compared. As a result, the tendencies of the features selected by each defect were clustered at the same point in the three-dimensional space and the classification performance of the gear defect was confirmed to be about 97%.