Flexible structure morphology reconstruction based on multi-core fiber grating sensors

Aerodynamic deformation of aerial refueling hoses under complex aerodynamic loads may lead to severe flight accidents, highlighting the significance of real-time perception of hose morphology. In this paper, we have developed a flexible hose morphology inversion method based on distributed multi-core fiber optic sensors. We propose the use of the RMF method, which transforms the multi-core fiber optic morphology reconstruction problem into a spatial curve inversion problem, addressing the issue of large reconstruction errors caused by the Frenet frame method's inability to define curvature and torsion at inflection points of spatial curves where curvature is zero. By employing multi-core fiber optic sensors, we achieve distributed sensing of multi-dimensional curvature information of the hose, resolving the complexity of the sensing network that arises from the need for single-core fibers to be laid out in multiple parallel paths along the hose surface. We have constructed a morphology monitoring and reconstruction system based on multi-core fiber optic sensors, with an average relative error of 4.96% in morphology reconstruction. The research outcomes are applicable to real-time inversion of flexible hose morphology in complex scenarios such as aerial refueling hoses and space-deployed trusses.