Developing Digital Twins of Dynamic Systems Using Vision Techniques, Multi-View Stitching, and Expansion Methods

Abstract

This study aims to create digital twins of structures such as wind turbines using noncontact techniques integrated with numerical methods. As part of this study, a stereo camera system mounted on a semi-autonomous drone captures the dynamic properties of a wind turbine. The drone with digital image correlation (DIC) cameras scans the entire structure to capture the deformation data of each field of view. The measured data includes the geometry, and displacement field data is mapped into a global coordinate system using 3D transformation matrices. The obtained data in the time domain for each field of view is transformed into the frequency domain to extract the operational deflection shapes and resonant frequencies for each field of view. The obtained deflection shapes are scaled and stitched in the frequency domain to extract the operating deflection shapes of the entire turbine. The operating data for the wind turbine is expanded using modal transformation techniques. A numerical model of the turbine is updated using expanded modal data. This methodology can be applied to any mechanical system subjected to dynamic loads. The work provides engineers with a new procedure to create and update digital twins using a limited set of measured data captured with non-contact techniques. The proposed approach to creating digital twins of structures can be used for full-field structural dynamic predictions, durability analysis, and structural health monitoring.