Open AccessCurrently, there is a tendency in the world that the unmanned aerial vehicles (UAVs) are beginning to be widely used in civilian areas. With the rapid development of the UAV, capable of moving in complicated terrain, the task of planning a real-time flight route is becoming more relevant and attractive.
Combining control methods of predictive models with and mixed integer linear programming can improve the efficiency of solving the problem of flight route planning in real time. In order to plan the optimal spatial trajectory of UAV when flying in difficult terrain (houses, mountains, etc.), in this paper, a novel approach to real-time three-dimensional trajectory planning for unmanned aerial vehicles (UAV) was represented under conditions of complex mountainous terrain, which can be built on the model of predictive control (MPC). Local terrain around UAV, which was modeled by triangulated irregular network (TIN) method as well as logical and continuous variables describing obstacle-avoidance are known within the limit detection radius.
However, taking into account the functional characteristics of the UAV, it is necessary to further treat smooth trajectory in its true time to receive the real-time permissible threedimensional trajectory. This article has been selected an algorithm for the serial connection of radius segments to smooth the planned route of flight of the UAV.
In the final part through the simulation results of the algorithm we have shown, using this algorithm, that the UAV successfully avoids all obstacles in real-time. This algorithm fully takes into account the limits on the maneuvering capabilities of the UAV, and it is proved that our algorithm is efficiently applied when the UAV moves in unknown environments, or in a situation of gradual obstacle detection in real flight.