Published: April 4, 2024
Language: Английский
Published: April 4, 2024
Language: Английский
Machines, Journal Year: 2025, Volume and Issue: 13(1), P. 60 - 60
Published: Jan. 16, 2025
Due to their precision, compact size, and high torque transfer, Rotate vector (RV) reducers are becoming more popular in industrial robots. However, repetitive operations varying speed conditions mean that these components prone mechanical failure. Therefore, it is important develop effective health monitoring (HM) strategies. Traditional approaches for HM, including those using vibration acoustic emission sensors, encounter such challenges as noise interference, data inconsistency, computational costs. Deep learning-based techniques, which use current electrical embedded within robots, address issues, offering a efficient solution. This research provides transfer learning (TL) models the HM of RV reducers, eliminate need train from scratch. Fine-tuning pre-trained architectures on operational three different conditions, healthy, faulty, faulty aged, improves fault classification across motion profiles variable conditions. Four TL models, EfficientNet, MobileNet, GoogleNet, ResNET50v2, considered. The accuracy generalization capabilities suggested were assessed diverse circumstances, low speed, fluctuations. Compared other proposed EfficientNet model showed most promising results, achieving testing an F1-score 98.33% each, makes best suited robotic reducers.
Language: Английский
Citations
0Welding in the World, Journal Year: 2025, Volume and Issue: unknown
Published: April 30, 2025
Language: Английский
Citations
0Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 192, P. 110250 - 110250
Published: May 5, 2025
Language: Английский
Citations
0Published: April 4, 2024
Language: Английский
Citations
0Published: April 4, 2024
Language: Английский
Citations
0