International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence (ICSMD2023)

84 / 2023-09-13 10:21:50

A parameterized optimization approach for gearbox dynamics response prediction

Parameterization,dynamics,gearbox,fault diagnosis

Special Sessions > SS15 - In-situ Inspection and Maintenance Robotics: Intelligent Sensing,Control, Locomotion, and Detection

Internal and full-field vibration testing of mechanical equipment is meaningful for equipment condition monitoring and fault diagnosis, but often cannot be achieved by typical vibration measurement techniques. As an alternative approach, vibration response prediction based on the dynamic model is helpful for addressing the problem but faces the challenge of insufficient accuracy in describing the system dynamics characteristics due to the existence of uncertain parameters. This paper proposes a parameterized implicit optimization approach for calculating the optimal values of uncertain parameters, thus enabling the prediction of dynamic responses. The experimental results demonstrate that the proposed approach can realize the reconstruction of the internal vibration with an error of less than 20% and can predict the main vibration characteristics of the gearbox to a large extent.