Digital Twin-Assisted Gearbox Dynamic Model Updating Toward Fault Diagnosis | Newswise

Research highlights that traditional virtual models for gearbox fault diagnosis often lack fidelity due to insufficient optimization against real-world data. This reliability gap limits the effectiveness of intelligent diagnosis, particularly when labeled training datasets are scarce or imbalanced. The study proposes a digital twin-assisted approach to dynamically update virtual models using physical-virtual data interaction. By leveraging parameter sensitivity analysis and optimization algorithms, this method aligns virtual signals more closely with physical realities, ensuring high-fidelity data generation. This advancement is crucial for open data initiatives because it democratizes access to high-quality, synthetic fault data. Reliable virtual datasets enable researchers worldwide to train robust diagnostic models without needing extensive physical testing resources, thereby accelerating innovation in industrial predictive maintenance while reducing the dependency on proprietary, real-world failure samples.

Source: newswise.com
Published on 2025-09-19

Digital Twin-Assisted Gearbox Dynamic Model Updating Toward Fault Diagnosis | Newswise

Related news