Planetary gearboxes (PGBs) are core industrial components, but current fault diagnosis faces high data collection costs, weak early fault signals masked by strong noise and low traditional model fidelity. This study aims to build a high-fidelity fault diagnosis model to address these issues and enable accurate PGB defect identification.
A rigid-flexible coupled multibody dynamic simulation (MBS) model integrated with the polygonal contact method (PCM) is constructed to simulate three PGB faults: sun gear broken tooth, bearing inner/outer ring defects. After experimental validation, Gaussian white noise mimics real conditions; quantum particle swarm optimization-optimized adaptive stochastic resonance (SR) enhances weak signals.
The model’s characteristic frequency relative error is < 4% (0.1% for bearing inner ring, 0.9% for outer ring). Under strong noise, adaptive SR amplifies bearing fault signal amplitudes by 10–15 times, enabling clear characteristic frequency identification and effective generation of practical PGB fault signals.
The model simulates simple fault scenarios without diverse noise types. Future research will incorporate multiple noise types and explore the correlation between fault frequency and operating time for PGB life prediction.
This low-cost, efficient solution enables accurate PGB defect detection in noisy environments, reducing maintenance costs and improving operational safety and reliability.
The study innovatively applies PCM to PGB bearing analysis, integrates MBS and adaptive SR to mitigate fault data scarcity, offering a reliable tool for PGB predictive maintenance.
The peer review history for this article is available at: Link to publonsLink to the website of publons
