Authors: Angel Rodriguez, Mohammad Shekaramiz, Mohammad A S Masoum, Abdennour Seibi
Mentors: Mohammad Shekaramiz
Insitution: Utah Valley University
The research was conducted using a Tello EDU drone, and this is a surrogate study that uses fan pedestals with removed blade cages to expose the blades, with one fan having damaged blades. Real-time object detection is done with Haar detection via Cascade Classifiers, while crack/anomaly detection is done with various deep learning classification models.
The proposed path planning and control framework for autonomous drones with low-cost positioning constructs a potential field based on the locations of the turbines if given. According to the potential field, simulated annealing is applied to the traveling salesman problem which generates a collision-free path for path tracking.
The proposed computer vision-based navigation system for autonomous flight of small unmanned aerial vehicles (UAVs) in GPS-denied constraints can detect cracks on wind turbine blade surfaces using drone-based inspection images. The deep learning model is trained on a large dataset of blade damages, collected by drone-based inspection, to correctly detect cracks.
The results of these papers can be used to improve the efficiency and safety of wind turbine inspections, which is crucial for the renewable energy industry as well as demonstrate educational aspects of autonomous systems to be taught in engineering courses.