Assembly state detection technology based on virtual reality for key components of large space deployable mechanism

Abstract

Aiming at the highly reliable assembly demand of large space deployable mechanisms with complex interface relationships of components such as bars and hinges, the intelligent assembly state detection of components based on Mixed Reality technology was studied. We designed and built the assembly condition detection system of key components of large space deployable mechanisms, realized the installation position detection of key parts with the Histogram of Oriented Gradients and Support Vector Machine as the recognition algorithm of truss joint base and image processing of screws with truss joint. The detection system is applied to the assembly of the large space truss mechanism. The results show that the accuracy of part recognition is 100%. The system can reduce problems of missing or the wrong assembly and improve the automatic level of large space deployable mechanism assembly detection.