The loosening of the bolts in the bolted connection structure will cause damage to the structure and even lead to disasters. In order to solve these problems, this paper combines the correlation analysis of structural vibration response under environmental excitation and deep learning technology based on stacked autoencoder (SAE), and proposes a bolt pre-tightening torque monitoring method based on structural vibration response. Firstly, using the time domain vibration response data of the structure under environmental excitation as the original information, the correlation index based on the inner product matrix method can be obtained to describe the structural health characteristics. Then, a stacked autoencoder for deep learning is constructed by the correlation index. Furthermore, the feasibility and effectiveness of the method are verified by experimental research on the bolt pre-tightening torque monitoring in the bolted connection plate. The results show that the method proposed in this paper has high accuracy in the monitoring of bolt pre-tightening torque.
Bolt Pre-tightening Force Monitoring Based on Stacked Autoencoder
Lect. Notes Electrical Eng.
Proceedings of the 5th China Aeronautical Science and Technology Conference ; Chapter : 33 ; 325-330
2021-11-02
6 pages
Article/Chapter (Book)
Electronic Resource
English
Bolt Pre-tightening Force Monitoring Based on Stacked Autoencoder
British Library Conference Proceedings | 2022
|Accurate bolt tightening by SPS
Automotive engineering | 1978
|Swedish Bolt Tightening System Adds Versatility
Automotive engineering | 1979
|Automatic bolt tightening - control principles and quality
Automotive engineering | 1987
|