Registration Number
0225U004810,
(0124U004665)
, R & D reports
Title
Intelligent system for recognizing defects in green energy facilities using UAVs
popup.stage_title
Розробити архітектуру розподіленої інтелектуальної системи розпізнавання дефектів об’єктів зеленої енергетики із використанням БПЛА
Head
Sachenko Anatolii O., д.т.н.
Registration Date
19-12-2025
Organization
Western Ukrainian National University
popup.description1
Improvement of methods for recognizing defects of green energy facilities using artificial neural networks based on video data collection and the use of BPLA
popup.description2
During the third reporting stage, core software modules for an intelligent UAV-based defect recognition system for green energy facilities were developed and verified. A fuzzy logic-based defect criticality assessment module, aligned with the IEC 61400–5 standard, was created. A dynamic trajectory adaptation subsystem was implemented to ensure UAV flight stabilization. A defect detection module based on a YOLOv8 neural network ensemble for multispectral data was developed, achieving a recognition accuracy of over 93 %. Additionally, algorithms for palette-invariant analysis and geo-spatial de-duplication were introduced for solar panel inspection, resulting in a 20 % reduction in false positives and a decrease in the volume of transmitted telemetry.
Product Description
popup.authors
Kochan Volodymyr V.
Zahorodnia Diana I.
Oleg S. Savenko
Zastavnyi Oleh M.
Oleksandr V. Melnychenko
Lesia O. Dubchak
Pavlo M. Radiuk
popup.nrat_date
2025-12-19
