A novel anti-loosening bolt looseness diagnosis of bolt connections using a vision-based technique

Abstract

Bolt looseness detection is a common problem in engineering. Most vision-based detection techniques focus on diagnosing ordinary bolt looseness, i.e., the methods used for diagnosis are based only on the sidelines of nuts. These methods cannot be used for anti-loosening bolt looseness diagnosis because of the simultaneous rotation of screws and nuts. Therefore, a novel anti-loosening bolt looseness diagnosis method based on a vision-based technique is proposed in this paper. First, a regular hexagonal cap was installed on the screw, which can be used as a reference for the nut. Then, to automatically distinguish the hexagonal borders of the screw cap and nut, a new hexagonal border reconstruction algorithm is proposed. Furthermore, the relative rotation angles of the screw cap and nut hexagons can be determined using the sidelines of the reconstructed hexagonal borders of the screw cap and nut. Finally, a novel anti-loosening bolt looseness diagnosis method is established by using the relative rotation angle of the regular hexagonal borders of the screw cap and nut under initial status and loose status. A prototype flange node of the transmission tower was used for experimental verification. The results show that the proposed method can effectively detect the loosening angle of anti-loosening bolts.