Damage Detection for Rotating Blades Using Digital Image Correlation with an AC-SURF Matching Algorithm

Abstract

The motion information of blades is a key reflection of the operation state of an entire wind turbine unit. However, the special structure and operation characteristics of rotating blades have become critical obstacles for existing contact vibration monitoring technology. Digital image correlation performs powerfully in non-contact, full-field measurements, and has increasingly become a popular method for solving the problem of rotating blade monitoring. Aiming at the problem of large-scale rotation matching for blades, this paper proposes a modified speeded-up robust features (SURF)-enhanced digital image correlation algorithm to extract the full-field deformation of blades. Combining an angle compensation (AC) strategy, the AC-SURF algorithm is developed to estimate the rotation angle. Then, an iterative process is presented to calculate the accurate rotation displacement. Subsequently, with reference to the initial state of rotation, the relative strain distribution caused by flaws is determined. Finally, the sensitivity of the strain is validated by comparing the three damage indicators including unbalanced rotational displacement, frequency change, and surface strain field. The performance of the proposed algorithm is verified by laboratory tests of blade damage detection and wind turbine model deformation monitoring. The study demonstrated that the proposed method provides an effective and robust solution for the operation status monitoring and damage detection of wind turbine blades. Furthermore, the strain-based damage detection algorithm is more advantageous in identifying cracks on rotating blades than one based on fluctuated displacement or frequency change.

A Drone-Driven X-Ray Image-Based Diagnosis of Wind Turbine Blades for Reliable Operation of Wind Turbine

“…In [9], the authors utilized RGB images obtained from drones to identify blade cracks by capturing Haar-like features for surface cracks on wind turbine blades. In [10], the authors proposed a modified speeded-up robust features (SURF)-enhanced digital image correlation algorithm combining an angle compensation strategy to detect cracks using the surface strain field of wind turbine blades. In [11], the authors proposed a deep learning-based automated damage inspection system to classify four damage classes based on a digital camera-equipped multi-copter UAV.…”

“…In [14], Alexnet, and in [15], [16], YOLO-based fault detection of wind turbine blade inspection has been done. In [10], the authors examined the operating structural characteristics of the rotating blades with digital image correlation. The proposed angle compensation (AC)-SURF algorithm detects and recognizes the relative strain distribution caused by flaws and any damage to the wind blades.…”

Acoustic-Signal-Based Damage Detection of Wind Turbine Blades—A Review

“…Digital image correlation (DIC) is another strain detection technique used for materials that calculates the deformation information of two digital images before and after damage, representing the displacement change in the region. In recent years, some newer DIC techniques have been proposed [15][16][17], such as a DIC technique using the AC-SURF matching algorithm for damage detection of rotating blades. However, DIC detection results are easily disturbed by occlusions.…”