Determining the direction of movement of the UAVs on the image of the underlying surface map

Problem statement. Unmanned aerial vehicles (UAVs), or drones, have gained popularity over the past decade. The use of autonomous drones appears to be a viable and low-cost solution to problems in many applications. Path planning capabilities are essential for autonomous control systems. An autonomous drone must be able to quickly calculate feasible and energy-efficient routes to avoid collisions. There are many challenges associated with developing a control system for autonomous drones. Several factors are needed for autonomous drone navigation, including decision-making capabilities, path planning, trajectory generation, and fail-safe backup control (in case of disturbances and failures). To achieve this, drones must have the ability to sense and perceive their environment and calculate a path using onboard computers and sensors. In addition, flight navigation controlled by an onboard computer reduces the frequency with which the operator needs to interact with the drone, thereby limiting the possibility of human error. Therefore, a method is needed to plan the shortest and most crash-free path for a UAV to an object in a dense urban environment. Objective. To propose a UAV path planning algorithm based on a sparse surface search algorithm for the shortest and safe driving in urban infrastructure. As well as a joint path planning algorithm for tracking a moving target in an urban environment using both UAVs and unmanned ground vehicles (UGVs). Results. The existing methods and algorithms for UAV trajectory planning are considered. During the comparative analysis, it was found that almost all path planning methods have shortcomings. Path planning algorithms for safe driving are developed based on a dynamic occupancy grid for modeling the target state and using a second-order Markov model to represent the target motion, joint use of UAVs and UGVs. The novelty of the work is that the algorithms take into account visibility impairments due to obstacles in the environment. The algorithm uses a dynamic occupancy grid to model the target state, which is updated based on sensor measurements using a Bayesian filter. The efficiency of the new methods is confirmed by the simulation results in environments with different building heights and different target speeds. The results of the proposed algorithm for different types of models, deterministic and Markov objectives are presented. Practical significance. The presented algorithm allows for trouble-free control of drones in dense urban areas based on the sparse surface search algorithm and dynamic occupancy grid.