A novel hybrid and publicly available model for spur gear vibrations based on an efficient dynamic model

Omri Matania, Lior Bachar, Roee Cohen, Jacob Bortman

Research output: Working paper/PreprintPreprint

Abstract

Dynamic models hold great potential for research and development in signal processing, machine learning, and digital twin algorithms for diagnosing rotating machinery. Various studies have suggested dynamic models of gears, employing many model approaches. However, there is currently a lack of a computationally efficient and publicly accessible model that accurately represents real-world data. In this study, we propose a novel hybrid model that integrates a realistic and efficiently validated dynamic model of spur gear vibrations with an enhancement process aimed at bridging the gap between simulation and reality. This process minimizes discrepancies between features extracted from simulated and measured data through fine-tuning of the model hyperparameters. The effectiveness of this hybrid model is demonstrated across numerous test apparatuses, encompassing several types of faults, severities, and speeds. The new hybrid model, inclusive of an upgraded dynamic model, generates data swiftly within seconds and is made publicly available with a user-friendly application programming interface and a detailed user manual. The novel suggested hybrid model has great potential to enhance future research in model-based studies, including machine learning, signal processing, and digital twin approaches. The user manual can be found in the following link: https://github.com/PHM-BGU/public_dynamic_model_for_gear_vibrations
Original languageEnglish
PublisherarXiv
Number of pages17
DOIs
StatePublished - 11 Oct 2024

Keywords

  • eess.SP

Fingerprint

Dive into the research topics of 'A novel hybrid and publicly available model for spur gear vibrations based on an efficient dynamic model'. Together they form a unique fingerprint.

Cite this