Research on gearbox fault diagnosis method based on EDT

Gears are crucial components in rotating machinery, and their health status serves as a significant indicator of whether machines can operate continuously. Therefore, an automatic and accurate method for early-stage gear fault diagnosis is essential in engineering practice. This paper proposes a new approach to gear fault diagnosis using the Euclidean Distance Technique (EDT). This method extracts the energy of frequency bands from vibration signals as characteristic parameters, and the Euclidean distance between the test sample and each status is calculated using a matrix that consists of six energy values. The effectiveness of this methodology is demonstrated through an experiment conducted on a real gearbox. Additionally, selecting band energy as a feature parameter proves to be more accurate than directly extracting signal energy.