Abstract
Drone applications have been developed extensively in different areas. For example, they are used in monitoring, searching, detection, communication, delivery, and transportation of cargo in various sectors of the economy. The reliability of drones in these problems should play a principal role. However the information for the reliability analysis can be incomplete and uncertain. This paper proposes a new method of drone fleet reliability analysis based on uncertain data. The mathematical model of a drone fleet is defined as structure function. The structure function is represented by a Binary Decision Diagram (BDD), which is constructed based on incomplete and uncertain data. The BDD construction is implemented based on the classification method of data mining, in particular, Fuzzy Decision Tree (FDT). This study shows that the transformation of inducted FDT to BDD is simplest than FDT transformation to other forms of structure function representation.