Abstract
This study presents clustering algorithms for the cleaning task of bird droppings on a solar panel using a foldable robot arm of a drone. Before the robot arm performs the cleaning task of bird droppings, a clustering task is required for a selective cleaning performance. In this study, two clustering algorithms were utilized: K-Means and self-organizing map (SOM). As the two algorithms exhibited different characteristics, appropriate strategies were required for cleaning different shapes of clusters to minimize the cleaning time. Accordingly, an efficient algorithm for selecting a suitable clustering method was proposed. The strategic algorithm was tested for several pattern examples on a solar panel.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have