Uncertainty evaluation for a drone-based laser triangulation system

The regular inspection of surface defects on rotor blades of wind turbines is crucial for a safe, reliable and efficient performance. However, manual inspections are laborious, require trained climbers and close proximity to the surface. Current automated systems operate from the ground, resulting in large measurement distances and thus, an insufficient resolution. To overcome the limitations of both manual and ground-based methods, a drone-based laser triangulation system is developed. The system is designed to inspect surfaces from a safety distance of approximately 1.5m up to 2m. As a result of using a difractive optical element, an imaging capability of a planar surface with 101 x 101 data points has been realized. Laboratory calibration of the designed laser triangulation system confirms a resolution of 0.4mm, which can be further improved using sub-pixel interpolation. Under rotor-induced vibration, the system demonstrates an expanded uncertainty of 1 mm at 95% confidence, accounting for rotor vibrations, sensor repeatability and sensor calibration. Thus, the results indicate the system's principle capability to detect surface defects in the sub-centimeter range with single-shot imaging.