Time-optimal trajectory generation for aerial coverage of urban building

Abstract

This paper presents a hierarchical architecture for generating the cooperative trajectories of multiple unmanned aerial vehicles (UAVs) attached with camera sensors, which aim to cover buildings with optimal time in 3D urban environment. It incorporates the centralized high-level layer performing the mission analysis and task allocation functions yielding instructions that are transmitted to the UAVs, as well as the decentralized low-level fashion that the UAVs perform the trajectory generation function in turn. First, the mission features especially the theoretical coverage time of building envelope are extracted quantitatively, and the buildings are then allocated to the UAVs in order to convert the cooperative control problem into multiple single-vehicle control problems. Then, each UAV visits and scans its allocated buildings sequentially, and the corresponding coverage trajectories are obtained by the parallel circle strategy (PCS) and the time-optimal guidance vector field (TOGVF) transition method, as well as the interfered fluid dynamic system (IFDS) method for obstacle avoidance. Finally, our proposed method is verified in various scenarios, and the simulation results show its high efficiency with least time to solve the cooperative coverage problem.