Алгоритм планирования маршрута БПЛА в условиях недостатка визуальных ориентиров

Abstract

Today, unmanned aircraft is used to solve a variety of tasks. One of the problems that arise when using unmanned aerial vehicles (UAVs) is the problem of navigation, in particular, determining their exact location in space. Today, there are various methods to solve this problem. A promising way to solve it in conditions where it is impossible to use a satellite navigation system is video navigation systems that use visual landmarks to determine location. In this paper, an overview of the existing types of UAV video navigation systems was carried out. The peculiarities of the operation of video navigation systems using the comparison of visual landmarks with satellite maps were also highlighted. The task was set to find a route in conditions of a lack of visual landmarks, in which the problem of losing reference points may arise. The graph data format used is described, which stores the values of the metric of the diversity of the underlying surface, which characterizes the magnitude of the probability of the problem of losing reference points. The existing algorithms for finding a path in a graph are considered and the need to develop a new algorithm is justified. A theoretical description of the algorithm was proposed, in which the path search process is divided into two stages: first, a solution option is searched for with the maximum value of the metric of diversity of the underlying surface, and then a route with the minimum length is selected. Several options of the implementation of the proposed algorithm were also presented and their asymptotic complexity was estimated. Based on the theoretical description, a software implementation of the proposed algorithm was developed and an experimental verification of the correctness of its operation on a test card with various flight tasks was performed. According to the test results, the proposed algorithm has shown its effectiveness by constructing routes correctly in all test examples. In the future, this algorithm can be used as part of the UAV video navigation system. The main advantages and disadvantages of the proposed algorithm were also described and ways for its further improvement were proposed.